Television Broadcasting Tape and Disc Recording Systems Ennes

M Tape and Dis
Recording System
®©
Television Broadcasting
Tape and Disc
Recording Systems
by
Harold E. Ennes
HOWARD W. SAMS & CO., INC.

THE BOBBS- MERRILL CO., INC.
INDIANAPOLIS KANSAS CITY NEW YORK
FIRST EDITION
FIRST PRINTING -1973
Copyright © 1973 by Howard W. Sams & Co., Inc., Indianapolis,

Indiana 46268. Printed in the United States of America.
All rights reserved. Reproduction or use, without express permis-

sion, of editorial or pictorial content, in any manner, is prohibited.
No patent liability is assumed with respect to the use of the infor-
mation contained herein.
International Standard Book Number: 0- 672 -20933 -0

Library of Congress Catalog Card Number: 72 -97059
Preface
Since the first television tape recorder was introduced in 1956, there
has been phenomenal growth in the use of video magnetic tape by telecast-
ers and independent production centers. It is difficult for the "old-timer"
to realize that many of his fellow workers do not remember the days before
tape, when practically two -thirds of the nation watched tv programs at odd
hours or from "hot kines" that had little resemblance to modern recordings.
Video tape, from the very start, provided superior picture quality, giving
a presence that cannot be matched by film. In addition, this medium pro-
vided the telecaster, for the first time, with a capability for local spot -com-
mercial and program production well within the means of a modest budget.
In spite of (or because of) the rapid growth, students and practicing
engineers have been faced with a scarcity of information relative to basic
recording systems. Having been privileged to attend training seminars at
both Ampex and RCA, as well as having had considerable practical experi-
ence with each system, the author has undertaken to fill this need. The pri-
mary purpose of this book is to provide fundamental knowledge for the
practicing engineer who feels the need for a beter understanding of his
equipment; this information may also serve as an introduction to the sub-
ject for the beginner.
Specific circuitry will undoubtedly change in the future as it has in the
past. Therefore, this coverage is general in nature, pointing up the primary
functions of video -tape equipment. The scope of coverage is from basic
theory to testing and maintenance of complete systems.
This text assumes a background in basic electronics theory. The reader
should also have a fundamental grasp of semiconductor logic circuitry, at
least of the same level as that presented in Chapter 3 of Television Broad-
casting: Systems Maintenance by Harold E. Ennes (Indianapolis, Howard
W. Sams & Co., Inc., 1972) . For the serious reader or student of broad-
casting, the following texts (by the same author and publisher) should be
considered prerequisites to this book: Workshop in Solid State, Television
PREFACE
Broadcasting: Equipment, Systems, and Operating Fundamentals, and Tele-

vision Broadcasting: Camera Chains.
The complete cooperation of the two major suppliers of television tape
recorders in the United States, Ampex and RCA, has made this book pos-
sible. Liberal use has been made of material supplied in factory training
programs of both manufacturers -in particular, descriptions of specific
circuit functions in a given system which have been integrated into the text
-as well as material gathered by the writer from various specialized
departments.
Credits to Ampex: For the "early days" of the VR -1000 and the VR-
1200, much credit is hereby extended to Larry Weiland for making plant
facilities available to the author, to Eldon Brown for his painstaking train-
ing seminar, to Tom Merson and Frank Gonzalez for their efficient sched-
uling of plant facilities tours, and to Joe Roizen for supplying much of the
requested material and photographs. For recent times ( the VR -2000B and
the AVR -1) , thanks are expressed to Frank B. Thompson, product man-
ager for the Ampex Audio -Video Systems Division, for supplying photo-
graphs and requested technical material, including the coverage of the
ACR -25 video cassette system.
Credits to RCA: For the "early days" of the TRT -1A and TRT -1B, ap-
preciation is expressed to John P. Taylor for supplying requested photo-
graphs and other material, to H. G. Wright for photos of the SMPTE tape
response, to Norman Hobson and Roy Marian for their answers to ques-
tions which confronted the writer in practical applications, to Ed Hill for
his most thorough scheduling of plant facilities inspection, and to John
Wentworth for his inimitable training sessions. For recent times (the
TR-70 series) , the thorough training sessions conducted by Bill Martin
were of highest value. Much credit is also due Dana Pratt, manager of
Southern broadcast sales for RCA, for supplying photographs and technical
coverage of the TCR -100 video cartridge system.
The author also extends his appreciation to the SMPTE and the Journal
of the SMPTE for permission to use material from SMPTE Recommended
Practice RP10 and to reproduce the paper "Electronic Editing of Magnetic
Television Tape Recording," by Norman F. Bounsall of Ampex, which
appeared in the February, 1962, issue of the Journal.
Also invaluable was the contribution of the Electronic Engineering Com-
pany of California (EECO) to the coverage of the latest electronic editing
systems.
HAROLD E. ENNES
Contents
CHAPTER 1
BASIC CONCEPTS IN QUADRUPLEX VIDEO TAPE SYSTEMS . 13
1 -1. The Time -Space Relationship 13

1 -2. Required Frequency Range for Television Tape 16
1 -3. Solving the Low -Frequency Problem 19
1 -4. Solving the High- Frequency Problem 21
1 -5. Recording Standards for Quadruplex Tape System 23
1 -6. Tape Systems 25
CHAPTER 2
THE QUADRUPLEX HEAD: THE BASIC TIME BASE . 27
2 -1. Orientation of Recorded Signals 27

2 -2. Head -to -Tape Velocity 30
2 -3. Tip- Penetration Effect on RF Level 33
2 -4. Timing 34
2 -5. Time -Space Errors 37
2 -6. The Headwheel Assembly 48
CHAPTER 3
BASIC REQUIREMENTS: WHAT THE SYSTEM MUST Do . . 57
3 -1. The Tape Transport 58

3 -2. The Recording Mode 60
CONTENTS
3 -3. The Playback Mode 65

3 -4. Synchronous Playback 69
3 -5. The Audio System 72
CHAPTER 4
THE TAPE TRANSPORT, AIR SYSTEMS, AND CONTROL CIRCUITRY . 76

4 -1. Functional Description of Tape Transport . 76
4 -2. Air Systems 78
4 -3. Transport Control Circuitry 85
CHAPTER 5
AND FM SYSTEM 89
5 -1. Functional Description 89
5 -2. Basics of the Modulation Process 92
5 -3. Demodulation Basics 95
5 -4. Refinements in Modern
Modulation -Demodulation Systems 102
5 -5. Modulation Circuitry 106
5 -6. Combined Modulation -Demodulation
Frequency Response 121
5 -7. The Video Preamplifier 125
5 -8. The Playback Amplifier 128
5 -9. Electronic Channel Switching 131
5 -10. Modern Demodulation Circuitry 145
CHAPTER 6
THE VIDEO PROCESSING SYSTEM 151

6-1. Functional Description 151
6 -2. Basic Principles of ATC 153
6-3. Monochrome ATC 158
6-4. Color ATC 162
6-5. The Tape Sync Processor 170
6-6. The Internal Reference Circuitry 172
CONTENTS
6 -7. Signal Processing 175

6 -8. Remote Video Gain 182
CHAPTER 7
THE SERVO SYSTEM 186

7 -1. Relationships 186
7 -2. Basics of the Ampex Intersync Servo System 189
7 -3. The RCA Headwheel Servo 201
7 -4. The RCA Capstan Servo 213
7 -5. The Control-Track Record and Simulplay Waveforms 223
7 -6. The RCA Headwheel Unlock (HW Unlock) Indicator 225
7 -7. The Vacuum-Guide Servo 229
CHAPTER 8
THE TAPE -DROPOUT COMPENSATION SYSTEM . 236

8 -1. The RCA DOC System 237
8 -2. The Ampex Dropout Compensator 249
CHAPTER 9
ADVANCED COLOR -ERROR CORRECTION SYSTEMS . 253

9-1. The Need for Advanced
Color-Error Correction Systems 253
9 -2. Basic Function 255
9-3. The Velocity -Error Corrector (VEC) 259
9-4. The Chroma Amplitude Corrector (CAC) . 271
CHAPTER 10
SPLICING AND EDITING FOR QUADRUPLEX RECORDING 283

10 -1. Mechanical Splicing 283
10 -2. Electronic Splicing (Ampex) 288
10-3. Electronic Splicing (1MX - 295
CONTENTS
10 -4. Principles of Programmed Electronic Editing . . . 309

10 -5. The EECO Editing System 314
CHAPTER 11
QUADRUPLEX TAPE SYSTEM OPERATIONS . 332

11 -1. Terminology 332
11 -2. The System Concept in Tape Operations 334
11 -3. Preliminary Setup Procedure 363
11 -4. Level Checks 364
11 -5. Head -Alignment and Self -Check Procedures 365
11 -6. Optimizing the Video Heads 375
11 -7. Checks During Recording 379
11 -8. Checks Before Playback 380
11 -9. Emergency Operations 381
11 -10. Electronic Splicing 387
CHAPTER 12
QUADRUPLEX TAPE SYSTEM MAINTENANCE 391

12 -1. Overall Performance Evaluation 391
12 -2. The Video -Head Assembly 394
12 -3. Optimizing the Local Sync Generator 396
12 -4. Checking Video Characteristics 398
12 -5. Use of Automatic Color Circuitry in Maintenance 419
12 -6. Checking Servo Stability 422
12 -7. Understanding Servo Functions 424
12 -8. Checking the Sample- and -Hold Circuitry . . 428
12 -9. Basic Applications and Servicing of Logic Circuitry 430
12 -10. Audio Systems Maintenance 441
CHAPTER 13
QUADRUPLEX -TAPE CASSETTE /CARTRIDGE SYSTEMS 451

13 -1. The Ampex ACR -25 451
13 -2. Ampex System Design Considerations 459
13 -3. The RCA TCR -100 468
CONTENTS
CHAPTER 14
VIDEO DISC RECORDING SYSTEMS . . . . 482

14 -1. The Single -Disc, Two -Head Video System . 482
14 -2. The Ampex HS -100 Video
Disc Recorder /Reproducer 484
14 -3. Basic Functional Description of HS -100 487
14 -4. The HS-100 Video Signal System 498
14 -5. Control Logic System 514
14 -6. The Ampex HS -200 Teleproduction System 539
APPENDIX A
TABLE OF PREFIXES 556
APPENDIX B
ANSWERS TO EXERCISES 557
INDEX 569
CHAPTER 1
Basic Concepts in Quadruplex

Video Tape Systems
Television tape recording requires a magnetic -field manipulation that

can adequately handle an 18- octave signal range at a satisfactory signal -to-
noise ratio. This chapter states the basic problems and the approaches to
their solutions.
1 -1. THE TIME -SPACE RELATIONSHIP
Although the time -space relationship is important to all communication

theory, television tape techniques tend to exaggerate the interdependence
of these quantities. The familiar frequency -to- wavelength conversion,
which is a fundamental time -space relationship, is given in Fig. 1 -1.
For any given medium, time and space have a fixed relationship. A given
medium indicates a fixed velocity, which is determined by the medium or
system in question. On waveform A of Fig. 1 -1, point 1 of a passing wave
is at reference point tl. One second later (waveform B) , point 1 of the
passing wave has advanced to t2, and point 2 is at the reference point. The
unit distance between tl and t2 is the wavelength, which also defines one
cycle of the waveform. One cycle per second is one hertz (Hz) .
The velocity of sound waves, although influenced by temperature,
humidity, and height above sea level, may be taken as approximately 1088
feet per second. Since the wavelength is equal to the velocity divided by the
frequency in cycles per second, a 1000 -Hz tone will have a wavelength (in
-
air) of 1088 1000, or 1.09 feet. Doubling the frequency to 2000 Hz
results in a wavelength just one -half as long. Increasing the velocity, how-
ever, increases the wavelength for a given frequency.
The tape velocity of broadcast -type audio tape recorders is either 71/2 or
15 inches per second (in /s). At 71/2 in /s, the recorded wavelength of a
1000 -Hz tone is 7.5/1000, or 0.0075 inch (7.5 mils, since one mil is 0.001
13
14 TELEVISION BROADCASTING
inch). At a tape speed of 15 in /s, the recorded wavelength of a 1000 -Hz

tone is 15/1000, or 0.015 inch (15 mils) .
The magnetic head gap must have a certain minimum physical size to
lay down adequate field strength on the tape. The relationship of head-gap
size to frequency is illustrated in Fig. 1 -2. As the frequency is increased
with the tape speed held constant, the recorded wavelength approaches the
physical size of the head gap. This results in zero output because of north-
south field cancellation of the signal. Therefore, the high -frequency limit of
tt t2
1 Sec -- --I
Fig. 1 -1. Fundamental time -space

relationship.
Velocity
the system is fixed by head -gap size and tape speed. Response can be ex-
panded into higher frequency regions by either or both of two methods:
(1) increased tape speed to produce a longer wavelength for a given high
frequency, (2) decreased gap size. See Fig. 1 -3.
Note that the higher tape speed extends the upper frequency response,
but, since the wavelength for a given low frequency is longer, the low-
Very Low Frequency, Small Rate of Change Across Gap
Medium Frequency
High Frequency With Wavelength Equal to Gap Size -

Zero Output
Magnetic Gap
Between Pole Pieces
Fig. 1 -2. Relationship between head -gap size and frequency.

BASIC CONCEPTS IN QUADRUPLEX VIDEO TAPE SYSTEMS 15
Higher Speed
m
inisi
yl
Fig. -3. Effects of head gap and
1
speed on frequency response.
I-
_r AI
f xs
SmallerIn
aim.
Noise Level
Frequency (Log Scale)
frequency response suffers. (There is a very small rate of change across

the gap; hence there is very small induced magnetic field energy.) Reduc-
Gap
ing the gap size means that less energy is available at the lowest frequen-
cies, so the low- frequency response again suffers.
The strength of the magnetic field is determined by the amplitude of
the signal fed to the head and the rate of change of the magnetic field.
For a given signal amplitude, as the frequency is increased (increased rate
of change) the magnetic -field intensity is increased. As the frequency is
doubled (increased by one octave) , the voltage is doubled. This voltage
increase, measured in decibels, is 6 dB. Therefore, a 6 -dB -per- octave rise
with increasing frequency occurs in magnetic recording. Fig. 1 -4A shows
the typical response (6 dB /octave) up to the high- frequency limitations,
where the response falls quite rapidly. Note that near the noise level and
below, the response cannot be called usable. The noise levels are tape
noise, modulation noise, and noise level of input amplifiers.
Conversely, as the frequency decreases from a given frequency limit, a
6 -dB -per- octave falloff in response occurs. At a certain low- frequency
limit, the change rate of the magnetic recording is so low, compared to the
gap size, that very little output exists. When the signal output falls to an
objectionable signal -to -noise ratio, the lower frequency limitations of the
system have been exceeded.
Fig. 1 -4B shows the typical equalization curve for correction of the mag-
netic recording characteristic of Fig. 1 -4A. The rising response at the end is
Response Limited by Losses
6 dB/Octave
CC z
Frequency (Log Scale) Frequency (Log Scale)
(A) Uncompensated response. (B) Typical equalization curve.

Fig. 1 -4. Frequency response in tape recording.
to correct for the rolloff at highest frequencies in the passband due to gap
size and tape magnetic- coating factors.
The frequency limitation of direct magnetic recording is about nine
octaves, or 30 to 15,000 Hz in the audio range. If it is desired to increase
the upper frequency limit to 30,000 Hz, the lower frequency limit must
also be doubled to maintain an adequate signal-to -noise ratio. The result is
still a nine -octave range, 60 to 30,000 Hz.
1 -2. REQUIRED FREQUENCY RANGE FOR TELEVISION TAPE

The range of frequencies required in a video signal extends to very low
frequencies for good picture shading and to comparatively high frequencies
for satisfactory fine detail in the picture. System requirements at the ex-
treme low- frequency end (approaching zero frequency, or dc) are aided
by line -to -line dampers and dc restorers; however, the low- frequency ac
response must also be very good. As a reference for this discussion, it will
be stated arbitrarily that 10 Hz is the low- frequency ac- response require-
ment. Response in this region is directly related to the proper operation of
clamps and dc restorers when these circuits are employed.
There are two resolution factors for a television picture -vertical resolu-
tion, which is independent of system bandwidth, and horizontal resolution,
which is directly related to system bandwidth.
Fig. 1 -5. Wedges for interpreting

vertical resolution.
The vertical resolution determines how well the horizontal lines in a pic-
ture are resolved. The maximum vertical resolution is fixed by the number
of active scanning lines. The United States standards call for a total of 525
lines per frame. Vertical- blanking time is approximately 7.5 percent of the
total frame time; therefore, the number of lines blanked out is 525 X 0.075
= 40 (approx) Since approximately 40 lines of the picture are blanked
.
out, approximately 485 active picture lines remain. This would appear to
indicate that 485 vertically spaced horizontal lines would be resolved, but
in practice this is not true. The slight spacing between the scanning lines
and the fact that the scanning spot straddles some of the lines both tend to
reduce the utilization of the maximum number of active lines. The reduc-
tion factor is usually considered to be 0.7 times the total active lines. Thus,
there are 485 X 0.7, or approximately 340 usable lines.
Fig. 1 -5 shows the horizontal wedges of a test pattern. Here, the lines
merge at a point that represents 340 black and white horizontal lines in the
total image height. This is a typical value of vertical resolution at both the
studio and transmitter outputs; hence, the television tape system should not
limit it.
Horizontal resolution is the ability to define vertical lines in the image.
The essentially round shape of the scanning spot and the fact that it is not
infinitely small place an immediate limitation on the ability to reproduce
rapid picture transitions. Thus, when the beam suddenly encounters the
sharp vertical line representing a transition from black to white ( Fig. 1 -6) ,
the resulting signal is not a square wave, but more nearly a sine wave.
Scanning Beam
--o- White
Black
Direction of Scan
Fig. 1 -6. Scanning beam encountering sharp vertical line.
Another factor contributing to the sine wave is the fact that a straight
vertical line represents an infinitely short rise time, which would require
infinite bandwidth. Such a bandwidth, of course, is impossible in practice;
the rise time of the signal representing the instantaneous transition is
limited by the available bandwidth of the system.
When the total of the rise and decay time equals the spacing between
lines, the lines are not visible as separate picture elements and are not
resolved. Rather, they appear as a solid area (Fig. 1 -7) . Assuming that the
scanning beam is properly focused, the limitation on horizontal resolution
is the system bandwidth. The pulse rise time representing an instantaneous
transition in the picture is directly related to the system bandwidth. As a
"rule of thumb," resolution of 80 tv lines requires a 1 -MHz bandwidth.
The gain-bandwidth product defines the relationship between a practical
amount of gain and the necessary bandwidth:
Fig. 1 -7. Wedges for interpreting

horizontal resolution.
f = Gain X Bandwidth
where,
f is the upper frequency limit for a gain of unity.
For example, for a given circuit with gain reduced to unity at 35 MHz:
Gain X Bandwidth = 35 MHz
Then for an uncompensated amplifier, the bandwidth for a gain of 10 is:
35 MHz
Bandwidth =
10
= 3.5 MHz
The relationship of bandwidth and rise time is of equal importance. The
product of bandwidth and rise time gives a factor, k, which depends on
type and magnitude of high- frequency compensation. If leading -edge over-
shoot is limited to less than 3 percent, factor k can be given a value of 0.35,
and the relationship between bandwidth and rise time can be expressed in
any of these three forms:
BW X RT = 0.35
0.35
BW =
RT
0.35
RT =
BW
Thus, the greater the bandwidth, the shorter is the rise time that can be
passed. For a bandwidth of 10 MHz:
RT =05
10
= 0.035 µs
Table 1 -1 relates bandwidth to rise time and horizontal -resolving power in

terms of TV lines.
The normal bandwidth of modern studio equipment is at least 8 MHz.
The picture transmitter, however, is limited to approximately 4 MHz by
Federal Communications Commission (FCC) channel assignments and
engineering regulations. Therefore, the horizontal resolution is restricted
to about 320 lines in the home receiver. The complexity of the problem
encountered in recording pictures on magnetic tape warranted a compro-
mise in this particular studio gear. Specifications for modern television tape
recorders include a bandwidth of at least 4 MHz (320 lines horizontal
resolution) with a signal -to -noise ratio of at least 35 dB. This meets the
minimum requirements for network- signal distribution as specified by
AT&T. These specifications are exceeded in high -band recording systems.
Table 1 -1. Relationship of Bandwidth, Rise Time, and

Horizontal -Resolving Power
Bandwidth (MHz) Rise Time (µs) TV Lines
1 0.35 80
2 0.175 160
3 0.1166 240
4 0.0875 320
5 0.07 400
6 0.058 480
7 0.05 560
8 0.0437 640
9 0.039 720
10 0.035 800
The video- signal frequency range for television tape extends from very
low frequencies (actually dc) to an upper frequency of at least 4 MHz. If
the range is considered to be 10 Hz to 4 MHz, a gamut of more than 18
octaves is required, which, as discussed earlier, is not possible in direct
magnetic -recording techniques.
1 -3. SOLVING THE LOW- FREQUENCY PROBLEM

The dc component in a standard video signal is inserted by means of
line -to -line dampers. Essentially, the clamper charges or discharges a
coupling capacitor to a dc reference, which usually represents the signal-
blanking level. This reference level assures that each active line starts from
the same reference immediately following horizontal blanking.
The extreme low-frequency requirements are met by employing an rf
carrier which is frequency- modulated by the video signal. (Frequency mod-
ulation was chosen over amplitude modulation to allow amplitude limiting
to be employed for attenuation of extraneous noise.) The carrier frequency
represents the dc component (either sync tip or blanking, depending on
the system clamp reference) Fig. 1 -8 shows the modulation characteristic
.
with a carrier frequency of 5 MHz clamped to the signal- blanking level.

The conventional assumption that the carrier frequency must be at least ten
times the highest modulating frequency is discarded in television -tape
applications.
IMPORTANT NOTE: The frequencies and deviation shown in Fig. 1 -8
apply only to low-band monochrome standards. The discussion in this

section and following Section 1 -4 apply to the same standard. Section
1 -5 will introduce the remaining standards now in use for quadruplex
recording systems.
With the carrier frequency clamped at the video-blanking level, it is

standard monochrome practice to adjust the picture- signal gain so that peak
white deviates 1.8 MHz upward in frequency; the peak -white signal occurs
at 6.8 MHz in Fig. 1 -8. With a standard video input, 0.3 -volt sync to 0.7-
volt video, sync tips deviate the carrier 0.7 MHz lower in frequency, to
4.3 MHz. The total deviation of 2.5 MHz is used currently as the 100 -per-
cent modulation reference for monochrome signals, in low-band video
tape systems.
When the modulation index is less than 0.5 ( frequency deviation less
than one -half the modulation frequency) , sideband energies beyond a single
pair are practically nonexistent, and the modulated signal approaches a -m
characteristics in this respect.1 Sidebands occur at the carrier frequency plus
At 2.5 MHz
/
\
1 MHz 4.3 MHz 5 MHz 6.8 MHz
/ \ /I 1
\\
I
Lowest sideband for Maximum Sync Tip I Blanking Peak

Video Frequency of 4 MHz I i White
I-0. 3V-r- 0.7 V -

Sync
Video
Fig. 1 -8. Frequency modulation of carrier for low -band monochrome recording.
and minus the instantaneous video frequency. The maximum video fre-
quency in the system passband determines the sideband limits. For a 4 -MHz
-
bandwidth, the lowest sideband occurs at 5 4, or 1 MHz. The upper side-
band extends just far enough to provide a shelf for the upper frequency
deviation. Sidebands beyond this limit are not used in low -band recording.
The only practical result of these compromises is a slight zig-zagging of
vertical lines apparent at a resolution of 300 lines and above.
The total frequency range now becomes 1 MHz to approximately 7 MHz;
thus, the modulation process has reduced the original 18- octave video range
to less than three octaves. A practical solution to magnetically recording
'Charles E. Anderson, "Signal Translation Through the Ampex Videotape

Recorder," Journal of the SMPTE, November 1958.
the video signal has been found if the problem of handling a 7 -MHz signal
can be met.
1 -4. SOLVING THE HIGH -FREQUENCY PROBLEM
A good audio tape recorder may have a head -gap size as small as 0.25 mil
(0.00025 inch) Then, for an upper audio limit of 15,000 Hz and a tape
.
speed of 7.5 in /s, the recorded wavelength is equal to:

7.5
Recorded wavelength =
15,000
= 0.0005 inch
= 0.5 mil
The recorded wavelength is twice as long as the gap width. This wave-
length /gap -size ratio is the lowest practical limit which can be adequately
pre- emphasized for good reproduction and signal -to-noise ratio. (Losses
in the magnetic core structure cause the signal to begin to decrease before
this point, and pre- emphasis is required.)
The smallest practical magnetic gap on the original video heads was
approximately 100 microinches, or 0.1 mil. Therefore, the minimum wave-
length for 7 MHz must be twice as long as 100 microinches, or 200 micro -
inches.
The required tape velocity can be calculated if the gap size and the mini-
mum required recorded wavelength at the upper frequency limit are known.
Since:
Video Record Video Playback
Inputs Outputs
Rotating Head Assembly
Record-Playback Relay
Commutator
Ground
II - II
Four Video Heads on
Drum Rotating At
14,400 rpm 1240 rpsl
Tape Pulled Toward

Observer at 15 inls
Fig. 1 -9. Parallel feed of four rotating heads (simplified diagram).

I -- --Af 2.5 MHz -I Af 3 MHz -- {
I I
I
I
dl
Mono I qÌ
Ir
I3
I
cÌ
I
I Mono I
m1
ml
5.5 5.79
I Id
I 6.5I
{ I
4. 28 5.0 I I 16.8 7. 06 7.9 10.0
i
I I i MHz
I
MHz
I I I I
Color Color (
I '
1
At1 MHz F- H
(A) Low-band. (B) High -band.

Fig. 1 -10. Frequency modulation in vtr systems.
Wavelength - Frequency
Velocity
then,
Velocity = Wavelength X Frequency

= 0.0002 X 7,000,000
= 1400 inches per second
If the tape were actually pulled across the head at a speed of 1400 in /s,
420,000 feet of tape on a reel over 80 feet in diameter would be required
System Modulation Sensitivity

13 MHz /Volt)
Peak White
50% Gray
1 Volt p-p
Picture Black 7.5% Setup
_ Blanking
Sync Tip
Applied Video
I
7.06 7.9 8.125 9.0 10.0
Carrier Frequency IMHzI
Fig. 1 -11. High -band recording standard for monochrome and color signals.
to record one hour of material. Try to visualize the required tape -transport
mechanism; obviously, this is an impractical solution.
The problem is solved by pulling the tape at a practical speed of 15 in /s
past a rotating video head. Insofar as the video signal is concerned, the
resultant velocity is more accurately termed head -to-tape velocity. The
rotating -head principle, illustrated in Fig. 1 -9, gives an effective head -to-
tape velocity of slightly over 1500 in /s. The fm video signal is fed to the
individual heads (quadruplex system) on the rotating drum and brush
assembly. Video tracks approximately 10 mils wide are laid down trans-
versely across the 2- inch -wide magnetic tape. Audio is recorded longi-
tudinally along the top of the tape at the conventional 15 -in /s rate. A 240 -
Hz control signal, which indicates the precise position of the rotating head
drum relative to the recorded video information, is recorded longitudinally
along the bottom of the tape. (NOTE: Details of connections between the
heads and the commutator are given in Chapter 2.)
1 -5. RECORDING STANDARDS FOR QUADRUPLEX

TAPE SYSTEM
The standard monochrome deviation for low -band systems shown in
Fig. 1 -8 is repeated at the top of Fig. 1 -10A. The sync -tip level is actually
at 4.28 MHz, generally rounded off to 4.3 MHz. For low -band color re-
cordings, the deviation is reduced as shown in the bottom portion of
Fig. 1 -10A.
Because of the strong color subcarrier component in the video signal,
interference and moire patterns result if standard monochrome deviation is
used for a color signal. Also, an excessive amount of differential phase and
gain of the color subcarrier results. Therefore, for low -band color recording,
the standard deviation is reduced to 1 MHz for 100 -percent modulation
(Fig. 1 -10A). Blanking is clamped at 5.79 MHz, the sync tip occurs at
5.5 MHz, and picture white peaks occur at 6.5 MHz. The main character-
istic that suffers here is a reduced signal -to -noise ratio.
Fig. 1 -10B shows the standard deviation for high -band recording sys-
tems, and Fig. 1 -11 is a graph of the high -band recording standard. Note
that the deviation is the same for monochrome and color signals. Blanking
level is clamped at 7.9 MHz. Sync tip occurs at 7.06 MHz, and peak picture
white occurs at 10 MHz. The deviation for 100 -percent modulation is then
-
approximately 10 7, or 3 MHz. NOTE: All tape recorders capable of
high -band operation are also capable of recording and playing back low -
band monochrome and color tapes.
Before recent advances, the head -gap size in video headwheels was 90
microinches. The gap size now has been reduced to 50 microinches. This
allows a higher frequency response as required for high -band recorders.
Improved design has resulted in as much sensitivity as existed with the
former (wider) gap. Also, due to more sophisticated design of filters and
pre- emphasis, de- emphasis circuitry, low -band performance is greatly

improved over the former low -band (only) recorder system. The per-
formance of modern high -band quadruplex recording systems can be
made so good that one is hard -pressed to judge whether the picture is from
a live or taped show. Where the high -band standard is used, even second -
and third-generation dubs can be made with almost negligible degradation
from the original performance. Almost all tape recording today is done on
the high -band standard.
VideolAudio
Master System Tape Waveform Control (Also
Erase Control Transport Video & Audio Monitor Picture Waveform Electronic Editor
Head Panel Panel Head Cover IA Scope) Monitor Monitor and Editec Panel
Editec Video& Intersync Electronic Editor

Colortec Control Panels Processing Amp
and Modules
Amtec Power Supply for
Video Signal System
Velocity Comp
Power Supply for
System Power Supply Intersync Servo System
Courtesy Ampex Corp.

Fig. 1 -12. Ampex VR -2000 video tape recorder.
1 -6. TAPE SYSTEMS
Modern television tape recorders of both Ampex and RCA design are
housed in a single console requiring no external rack space. External racks
normally house video and audio signal- source selection panels, color moni-
tors, and vectorscopes.
A typical tape "system" is the Ampex VR -2000 video tape recorder
shown in Fig. 1 -12. The rotating video head drum and all audio heads are
under the protective head cover. The front of the lower section houses all
system modules, and the rear of the lower section contains the required air
systems and intermodule wiring harness.
Monitor
Deck
Secondary
Control Panels
(Cover Open)
Transport
(Cover Open
Primary
¡- Control Panel
Electronics Air Electronics

Bay 1 System Bay 2

Fig. 1 -13. Ampex AVR -1 tape system.
Normal system inputs are ac power, selected video and audio signal
sources, composite sync, and 3.58 -MHz subcarrier.
Fig. 1 -13 illustrates the Ampex AVR -1 tape system. This is a complete
and automatic system itself, or it may be used on a "shared electronics"
basis for the Ampex ACR -25 video cassette system (Chapter 13) .
EXERCISES
Q1 -1. From your studies thus far, and your understanding of Figs. 1 -3 and 1 -4,
would you conclude that the low video frequencies need to be brought
up in amplitude relative to higher video frequencies?
Q1 -2. If you have an amplifier with unity gain at 50 MHz, what is the usable
bandwidth at a gain of 5 (uncompensated) ?
Q1 -3. Assume that the head -to -tape velocity is 1560 inches per second. What
is the recorded wavelength of a 10 -MHz signal?
Q1 -4. What is the shortest usable wavelength on the tape for a head gap of
50 microinches?
Q1 -5. What frequency corresponds to your answer for Q1 -4, assuming the
same velocity as in Q1 -3?
Q1 -6. What is the term applied to the sideband above the highest deviation
frequency in video tape recording?
Q1 -7. In the drawing of Fig. 1 -8, a standard 1 -volt p -p composite signal input
is assumed. Give the sync -tip frequency and peak -white frequency re-
sulting from an input which overshoots to 1.5 volts p -p (video -to -sync
ratio maintained and no video clipping or limiting used in recorder).
CHAPTER 2
The Quadruplex Head:

the Basic Time Base
You will now learn the precision of the four-head (quadruplex) assem-
bly. By acquiring the time -base concept, you will more clearly understand
the relationship of the recording system to the television signal itself.
You will learn later that as precise as the rotating head assembly is, more
than one -half of all the rest of the system function is concerned with cor-
recting for deficiencies of this assembly. This is the reason why the time -
base concept is so important.
The circular mounting which contains the four video heads is termed the
headwheel by RCA, and the drum by Ampex.
2 -1. ORIENTATION OF RECORDED SIGNALS

A rotating head contacts the tape for an arc of approximately 120 °.
Since the head rotates at 240 rps, it takes 1/240 second for 360 °, and one -
third of this time (1/720 second) to traverse the 2 -inch tape ( 120 °) .
Inasmuch as the tape travels at a rate of 15 inches per second, it will go
15/720, or 0.02, inch (20 mils) while the head describes its arc across the
tape. Therefore the bottom of each video track is ended 20 mils later than
the start, resulting in an angle of 0.54° (Fig. 2 -1).
The orientation of recorded information on a vertical tape transport is

shown in Fig. 2 -2A. The video heads contact the entire 2 -inch surface of
the tape; however, a 70 -mil -wide area across the top is erased for audio
information, which is recorded on a track parallel to the edge of the tape.
The video tracks along the bottom of the tape, containing the control and
cue track, are not actually erased; however, only that portion used as video
information is shown in Fig. 2 -2A.
Tapes recorded on horizontal and vertical transports are entirely com-
patible and may be played back on either type of transport. The track
27
Start
0.54° lar 33 Minutes)
tan 00. 02 -0.01
then 0 0.54°
Track Slant
End
Tape Travel at 15 in /s
Fig. 2 -1. Track slant on tape.
Video -Head Motion -ar

Headwheel Motor 10 Mils Video
5 Mils Guard
AHeadwheel Video
Tape
Travel Sound Track - Bottom of Taoe
F- Control Track
145 Mils)
Top of Tape
Top View
Tape Travel
Cue Trac k
70 Mils- [-
1.82-
Looking Toward Tape
From Headwheel Side 2"
(A) Vertical transport.
Sound
Tape Travel Track Top
Drum Motor Drum

_/
Tape Video -Head Motion

Travel
Looking Toward Tape

From Headwheel Side
Cue Control
Track
(B) Horizontal transport.

Fig. 2 -2. Track orientation.
THE QUADRUPLEX HEAD: THE BASIC TIME BASE 29
Air
/,
'
'
sure
Vacuum Guid
r
(A) Top view.

/
A
Head
Tip
Penetration
V
.4
Tape
Fig. 2 -3. Cross -sectional view of vacuum guide.

(B) Side view.
orientation on the tape is that which is viewed looking toward the coated
side from the rotating heads. The track dimensions on the horizontal
version shown in Fig. 2 -2B are exactly the same as for the vertical arrange-
ment in Fig. 2 -2A.
The method of holding the tape concentric with the rotating heads is
shown in Fig. 2 -3A. The vacuum guide holds the tape in place, and the
center slot provides clearance for tape stretch under head penetration.
Fig. 2 -3B is a side view of the vacuum guide and the rotating headwheel.
Most modern tape systems provide for a dual -speed tape transport (71/z
or 15 in /s). Dual -speed operation employs standard quadruplex recording
techniques, using standard two -inch recording tape. All the advantages of
quadruplex recording are retained, and the cue track, audio track, and
control track are untouched. Only the method by which video tracks are
laid down is modified.
The headwheel employs four video heads which record a track only 5
mils wide, as compared with 10 mils for conventional recording. Fig. 2 -4A
shows how the 5 -mil tracks at half speed are spaced with respect to the
5-Mil 5 -Mil 10 -Mil

1111111111111111111111111111111
Track Track Track
1111111111111111111111111111111
2.5 -Mil
1111111111111111111111111111111
10 -Mil 5-Mil
Spacing 1111111111111111111111111111111 Spacing Spacing
(A) Half-track, (B) Half-track, (C) Full -track,

71/ in/s. 15 in /s. 15 in /s.
Fig. 2 -4. Dual -speed recording tracks.
video portion of the tape (5 -mil track with 21/2 -mil spacing) . The capstan
motor pulls the tape at 71/2 in /s; thus, twice as much video information is
packed into each foot of tape. Because the speed of the headwheel rotation
is the same as in normal operation, each recorded track contains the same
number of tv lines, approximately 16.5.
During operation at 15 in /s, video tracks are recorded by the half -track
heads as shown in Fig. 2 -4B. In this case, the track width remains at 5 mils,
but the spacing between tracks becomes approximately 10 mils. The tracks
in Fig. 2 -4B correspond to every other track in Fig. 2 -4A.
Standard video -headwheel panels recording 10 -mile tracks with approx-
imately 5 -mil spacing between them (Fig. 2 -4C) are oriented in the same
manner as in Fig. 2 -4B. Comparison of Figs. 2 -4B and 2 -4C will show how
interchangeability is achieved.
If a 10 -mil head traces a 5 -mil track, the head is in contact with the
signal included on the track, and, in addition, it traces over 5 mils of blank
tape. It encounters the next track after a 5 -mil spacing interval which
brings the head into position to read all the information on the next track.
In the case of a half -track (5 -mil) head reading a fully recorded (10-
mil) track, the principle is reversed. Here, the half -track head reads enough
of the video information (5 mils) on the fully recorded track to produce
high -quality pictures.
2 -2. HEAD -TO -TAPE VELOCITY

The wheel which contains the four video heads has a reference diameter
of 2.064 inches. The tip projection of a new head is about 3.6 mils, and for
a worn head it is about 1.0 mil (Fig. 2 -5) .
The maximum recording diameter (tip projection of 3.6 mils) is:

A +B +C= 0.0036
2.0640
0.0036
2.0712 inches
The recording circumference is:
Circumference = lyd
= 3.1416 X 2.0712
= 6.507 inches
Since the head revolves at 240 rps, the head -to-tape velocity at the maxi-
mum tip projection is:
Head -to -tape velocity = 6.507 X 240

= 1561 in /s
A B C
I ¡
Tip Projection A & C - 3.6 Mils Max

I.OMiI Min
Ref Diameter B = 2.064"
Recording Circumference
Fig. 2 -5. Headwheel reference diameter and tip projection.
The recording diameter at the minimum tip projection of 1.0 mil is:
A +B +C =0.001
2.064
0.001
2.066 inches
The recording circumference is:
Circumference = and
= 3.1416X 2.066
= 6.490 inches
The head-to -tape velocity is:
6.490 X 240 1557 in /s
at minimum tip projection of 1.0 mil.
HEAD -TO-TAPE VELOCITY AT 3.6 -mil TIP PROJECTION = 1561 in /s.

HEAD-TO -TAPE VELOCITY AT 1.0 -mil TIP PROJECTION = 1557 in /s.
The foregoing relationship of tip velocities has a direct bearing on

proper timing to avoid space errors in television -tape playback.
The angular velocity remains the same regardless of tip projection. The
spacing of the horizontal -sync pulses across the entire video track must also
remain the same to avoid time -space errors in the reproduced picture.
That is, the space occupied by a tv line must be the same at the middle and
end of the video track as it is at the beginning of the track.
Fig. 2 -6A shows the tape without the head engaged. As the vacuum
guide is moved toward the head, the pole pieces contact the tape and a
stretch occurs as shown in Fig. 2 -6B. The greater the tip penetration, the
greater is the stretch and the wider is the space between sync pulses. The
effective spacing in time between pulses is determined by the tip velocity.
A recording made with a 3 -mil penetration can be played back with a worn
head having a tip penetration of only 1 mil. This is because the reduced
tip velocity of the worn head means a longer time is required to scan a
given length of the video track; therefore less tape stretch is required to
maintain the correct space-time relationship between pulses.
Pulses
Farther
Apart
Equally Spaced
Sync Pulses
Drum for Headwheel)
Pulses Equally Spaced
(A) Tape without head engaged. (B) Tape with head engaged.
Fig. 2 -6. Pole -tip stretch of tape.
From this analysis, it is evident that reduced velocity due to worn tips
is complementary to the resulting reduction in tape stretch. Therefore,
assuming that the video -tape stock neither shrinks nor stretches during its
life, the vacuum -guide position relative to the headwheel remains the same
as tip projection decreases with head wear.
To emphasize this point, see Fig. 2 -7. A new head with a 3.6 -mil tip
stretches the tape considerably more than when this same head is worn
down to a 1 -mil tip. But note that the head -to -tape velocity is 4 in /s
greater for a 3.6 -mil projection than when the head is worn to a 1 -mil
projection.
This is not a geometric (space -in-time) error if the vacuum guide has
not been moved, since the standard radius of rotation is maintained. In
Fig. 2 -7B, the arrow corresponding to maximum tape stretch represents
the length of the pole -tip contact with the tape, which is about 16 video
lines. The other arrow indicates the shorter length of the same 16 lines
under minimum stretch, but the time taken to scan the same space is
longer; hence there is no geometric error.
Keep the time -space concept. Errors in timing ( velocity errors) result
in horizontal displacement of vertical lines in the picture (Section 2 -5) .
Velocity errors result when the time to scan a given space (such as the
space occupied by 16 lines) differs from the timt elapsed in recording
that space on the tape.
Radius of Rotation Between 1.0329 and 1.0356 Inches
Effective Diameter 2.0658 to 2.0712 Inches
Tape at Maximum Stretch
Difference in Stretch
Pole Tip @ 1 Mi
Pole Tip @ 3.6 Mil
Tape at Minimum Stretch
Drum
Minimum St etch 6
Length Maximum Stretch
1 Length
@ @ ' ,
1561Ints 1557 In /s
Fig. 2 -7. Change in tip projection does not cause: geometric error.
J
A nonstandard recording radius can be used to -flake a recording on a
tape. If the tape is played back on the same head with the same adjustment,
no geometric error results. This simply means that the playback space-in-
time velocity is the same as the recorded signal space -in -time velocity.
For interchangeability of tapes and to facilitate splicing of sections from
different types, it is necessary to adhere to a standard pole -tip arc radius,
within very narrow limits. Even when this standard Tolerance is maintained,
velocity errors can result from slightly different guide-height adjustments,
tolerances in tape tensions between machines, slightiy different positions of
tape-transport fixtures, and tape stretch resulting from use. (These prob-
lems are expanded in Section 2-5.)
2 -3. TIP -PENETRATION EFFECT ON RF LEVEL
The position of the vacuum guide, which cups the tape around the
rotating heads, determines the amount of pole -tip penetration into the
tape. When the guide is sufficiently far away from the pole tips, a positive
clearance between tape and heads occurs, and no contact exists. Under this
condition, no recording would be laid down on the tape, and no playback
of recorded information can occur (Fig. 2 -8A) .
Radio -frequency output from recorded information begins to occur as

the vacuum guide is brought closer to the head. There is now a negative
clearance between the pole tips and the tape in the slotted guide. The
amount of this negative clearance is the tip penetration. When the penetra-
tion is extremely light, concentricity between the tape and head does not
exist (discussed further in Section 2 -5 ). Under this condition, tip penetra-
tion is less at the center of the tape than at either edge, resulting in a dip
(Fig. 2 -8B) in the rf envelope of the head output. As the tip penetration
is further increased, intimate contact is gained with full concentricity, and
an even rf output (Fig. 2 -8C) is obtained from the head.
(A) No contact, zero (B) Penetration too (C) Full contact,

output from tape. light, dip in center even rf output.
of envelope.
Fig. 2 -8. Relationship of rf output and pole -tip contact with tape.
An adequate magnetic contact must prevail at the low values of tip pene-
tration which exist near the end of head life. Therefore, since the guide
position relative to the head is a constant value (Section 2 -2) , new heads
with high tip projection will exhibit relatively high values of tip
penetration.
High values of tip penetration increase head and tape wear and the load
on the head-motor drive amplifiers. If excessive loading occurs, head -servo
instability can occur. This is particularly true when marginal tubes are in
service in the older tube -type recorders.
A tip penetration of less than 1 mil aggravates dropouts and possible
head clogging. Dropouts are white flashes in the reproduced picture; they
are caused by microscopic irregularities in the magnetic coating of the tape.
Head clogging results when iron -oxide particles, loosened from the tape
coating, become lodged in the head gap. (See Chapter 8 for tape-dropout
compensation systems.)
2 -4. TIMING
The headwheel (or drum) rotates at 240 rps, or four times the tv -field
frequency of 60 Hz. Thus, one revolution (360°) of the wheel lays down
four tracks (four heads spaced 90° and fed simultaneously) which are
equal to one -fourth of a field. Since there are 262.". lines per field, there
are 65.625 lines in four tracks on the tape, and each track contains 16.4
lines.
NOTE: A tv line is designated as H, where H is the interval from the
leading edge of one horizontal -sync pulse to the leading edge of the next
horizontal -sync pulse. (The standard interval is 63.5 microseconds.)
Therefore, a recorded track equals 16.4H.
The preceding is the average number of picture lines used in repro-

ducing the signal; actually, each track contains more than this number of
lines. The guard bands and control, cue, and audio tracks (Fig. 2 -2) leave
approximately 102° of the tip pass for picture information. The number
of lines per degree of rotation can be calculated as Follows. First, find the
time for 90° of rotation:
Time per revolution = second
= 4167 microseconds
Time for 90° =4 4 = 1042 microseconds
Since the duration of one line is 63.5 microseconds, the number of lines
in 90° of rotation is:
1042
63.5
- 16.4 lines
Then the number of degrees per line is:
90 = 5.48°
16.4
and the number of lines per degree is:
1 = 0.182 line
5.48
The heads are 90° apart and lay down 16.4 tv lines in 90° of rotation.
Since the heads are spaced 90 °, the remaining time that both heads are in
contact with the tape is 102 ° -90 °, or 12 °. This leaves an overlap of
0.182 X 12, or 2.2 lines (Fig. 2 -9) In recording, the same information is
.
duplicated in these 2.2 lines. During playback, electronic switching is used

to disconnect the head nearing the end of a track and to connect the head
beginning the next track. The switching is done during horizontal retrace
to avoid visible switching transients.
During 240 revolutions of the headwheel, the tape travels 15 inches.
One revolution of the headwheel (four tracks) is equal to one -fourth of
Useful Picture Information

= 102° lapproxl on Tape Head 2 Enters Tape
°
102 Before Head 1 Leaves
By 102 -90 12°
Fig. 2 -9. Active picture information and overlap on tape.
Table 2 -1. Significant Head Timing
Headwheel
Rela- Degrees
Live Vertical 360° =
Fre- Micro- (V = 1/60 TV Lines 1/240 sec
quency seconds sec = (H = = 4166.6
Subject (Hz) (µs) 16,667 As) 63.5 As) As
TV Field 60 16,667 1 262.5 1440

TV Line 15,750 63.5 1/262.5 1 5.48
1 Revolution of
Headwheel 240 4166.6 1/4 65.625 360
1/2 Revolution of
Headwheel 480 2083.3 1/8 32.8 180
1/4 Revolution of
Headwheel 960 1041.6 1/16 16.4 90
Playback Switching
Rate (and
Sweeps Across
Tape per Sec) 960 1041.6 1/16 16.4 90
1° of Rotation 11.58 0.182
Additional Significant Measurements

Width of video -head magnetic gap 0 05 mil
Video -head tip speed in /s 1561
Length of one tv line on video track 98 mils (approx)
Head -to -head spacing around circumference 1 626 inches (approx)
Length of video track across tape 82 inches
1
Space along tape between fields 0 25 inch

Space along tape between frames 0 50 inch
-
(240 4) recorded on 15 inches of
a field. Therefore, there are 60 fields
tape. Thus one field equals 0.25 inch (15 _ 60) , and a frame (2 fields)
equals 0.5 inch of tape.
The foregoing timing relationships and other significant factors are
summarized in Table 2 -1.
NOTE: The tracks on video tape can be made visible by coating the tape
with a suspension of carbonyl iron and diluent (Fig. 2 -10) Horizontal - .
sync pulses show as white dots between the long vrhite lines of vertical
blanking. Fig. 2 -11 shows how the picture raster is made up of consecu-
tive bands of 16 to 17 lines contained in each head pass across the tape.
Visible tracks such as those in Fig. 2 -10 are used only for mechanical
splicing. This method has largely been replaced ly electronic splicing
and editing systems (Chapter 10) .
White marks in this control -track area are frame edit pulses.
112"
1 Frame
a'
1 Field

Fig. 2 -10. Visible tracks on tape for mechanical splicing.
2 -5. TIME -SPACE ERRORS
Errors in spacing, as a function of time, in a video -tape reproduction

occur as horizontal displacements of vertical lines i the picture. We will
now study the space errors caused in the reproduced picture by time -base
discontinuities.
A headwheel assembly consists of the rotating headwheel and the vacuum
guide (details and illustrations in Section 2 -6) . For a given headwheel
assembly, the vacuum guide has a radius of from 1.0329 inches minimum
to 1.0334 inches maximum. The rotating headwheel has a pole -tip arc
radius of 1.0329 inches minimum to 1.0356 inches maximum (Fig. 2 -12).
A nominal tape thickness of 1.4 mils is assumed.
See Fig. 2 -13A. If the guide position is adjusted so that the center of
curvature of the guide and the center of rotation of the headwheel exactly
coincide, then precise concentricity is achieved. This means that the head-
to -tape pressure remains exactly the same across the entire arc of head-to-
tape contact. Under the condition of Fig. 2 -13A, playback of a properly
Raster is constructed from 16 -17 lines (average 16.41

Horiz Blanking in each of 16 bands at a 60 -Hz repetition rate. Head No
2- Vert Blanking Ends

3
16 -17 Lines Each Head 1
I
2
4
1
Start of Vert
Blanking and Sync
Done During Horiz Blanking Interval
Fig. 2-11. Raster makeup, with vertical sync on head 1.
recorded standard tape will yield a picture that, when viewed without any
automatic time correction (ATC) or other time -base correction (Chapter
6) , is free of geometric errors.
If the guide is now moved away from the headwheel as in Fig. 2 -13B, we
have the condition already discussed relative to Fig. 2 -8B. A high head -to-
tape pressure occurs at the top and bottom of the head pass, and a low
Vacuum guide moves in or out

to obtain proper engagement
with rotating head pole tips.
I
A
A Vacuum Guide Radius: 1.0329 Inches Minimum

1.0334 Inches Maximum
B Radius of Pole -Tip Arc: 1.0329 Inches Minimum
1.0356 Inches Maximum Reference: SMPTE RP11
Fig. 2 -12. Guide and pole -tip radii.

Vacuum Guide
(Moves In and Out)
Same Head-to -Tape

Pressure Across
Entire Arc
\
Radius of Pole
Radau
Tip Rotation
Fixed Center Rotatirg Wheel ÍI
(A) Guide and headwheel precisely concentric.
Original Concentric Arc IN'w+ Lost

Vacuum Guide
(Moved Away From Head)
High
Head -to -Tape
Pressure
//
Radius of Pale
,Tip Rotati:n
\ \
l'l
Low
Head to Tape
Pressure
- i--I
I
Original
Concentric
Center
--'y- Fit
Moved Away
Center of
Relating Wheel
1
'
I
I
\
1\\ /
High
Head -to -Tape
Pressure \\ ././///
(B) Guide moved away from headwheel.

Fig. 2 -13. Relative positions of guide and headwheel.
pressure occurs at the center. Note that a measure of actual tip projection
of the headwheel is that obtained when the contact at the center of the tape
is lost under nonconcentric (eccentric) operation. We may now study the
various forms of geometric distortion, starting with the example just cited,
the vacuum guide adjusted too far from the headwheel.
Tape Vacuum Guide Too Far From Head
Point Z in Fig. 2 -14A represents the center of the drum, or headwheel.
Assume that the recording is made with arc X -X representing 16 lines.
Y
Enters Late
X -X - 16 Lines on Tape
Y -Y 16 Lines on Tape
X-X -Z = Recording Dimension
Leaves Early\ Y -Y -Z = Playback Dimension
Y(
(A) Recording and playback dimensions.
Starts Late ///////////%/,.

Ends Early
16 Lines llllllllllllllllr.
16 Lines
lllllllllllllllll.
/ / / / / / / / / / / / / / / / /
16 Lines
lllllllllllllllllll
/ / / / / / / / / / / / / / / / // (
lllllllllllllllllll
/
Lagging Skew
/ / /
/ / / / / / / / / / / / / / /l
ps Error
I
i
'llllllllllllllllll
/ / /
/ / / / /
lllllllllllllllll
/ / / / / / / / // /
Peak -to -peakerror in ps is equal

to guide error in mils lapproxl. /
(B) Bands of lines. IC) Effect on vertical l nes.
Fig. 2 -14. Space distortion, vacuum guide too far from head.
Arc Y-Y is the playback position, if the guide is too far from the heads,
and contains the same 16 lines of information. The pole tip enters this
information late and leaves early compared to the recorded time base.
Space errors resulting from the condition in the preceding paragraph
are depicted in Fig. 2 -14B. Each band of lines representing a head pass
across the tape starts late and ends early, producing the skew effect, also
called the venetian-blind effect.
Fig. 2 -14C shows the effect on a series of vertical lines (viewed on a

video monitor) when time -base distortion, as a result of having the
vacuum guide too far from the head, is present.
NOTE: The horizontal position of the vacuum guice (which determines

the distance of the tape from the heads) is controlled by a servo amplifier.
This servo (in the playback mode) may be placed either in the manual or
automatic sense position.
In practice, mechanical adjustments are made on the head assembly with

the electrical controls centered. After the initial mechanical alignment of a
given head assembly, only the electrical control need be used to eliminate
skew in the manual mode of operation (Chapter 11) Details of the guide
.
servo in the manual and automatic modes are found in Chapter 7.
Interchange S- Distortion
Note from Figs. 2 -14B and 2 -14C that the skewed lines are not straight
diagonal lines, but that each segment is a 90° section of a sinusoid. A
single head scan across the tape is 90° of the headwheel rotation rather
than the full 360 °, so only 90° of the sinusoid occurs on each band of 16
to 17 lines.
Note also from Fig. 2 -14B that the number of microseconds of error
(peak-to -peak) on vertical lines in the picture is approximately in direct
ratio to the guide error (tip -projection error) in mils. Thus if the guide is
misadjusted by 0.5 mil, a timing error of approximately 0.5 µs occurs.
If the guide is misadjusted 1 mil, an error of approximately 1 µs results.
This error is approximate because a 1 -mil misad¡ustment of the guide
results in exactly 0.92 Its of timing error rather than 1 µs, but the rule of
thumb is to use a direct microsecond -to -mil relationship. This information
is useful in judging the amount of tips remaining on a headwheel, as
outlined in Chapter 11.
Note from Fig. 2 -12 that the tolerance on the guide radius is from
1.0334 inches maximum to 1.0329 inches minimurr. Thus the tolerance is
-
1.0334 1.0329, or 0.0005 inch (0.5 mil) Let us see what this seemingly
.
tight tolerance actually means in practice.

Assume that a "standard recording" (precisely concentric guide and
pole -tip arc) has been made on a headwheel assembly with the maximum
guide radius. Further assume that this same tape is now played back on a
headwheel assembly which has the minimum guide radius (0.5 mil smal-
ler). The panel is adjusted for exact concentricity. Fig. 2 -15A shows the
result. When precisely concentric operation is carried out, the smaller
guide radius gives a geometric error (timing error) which is 0.5 µs in the
skew direction, indicating that the guide to too ca'ose to the headwheel.
Note, however, that this skew error consists of lines that are perfectly
straight.
Now if this guide is backed away from the headwheel assembly 0.5 mil,
concentric operation is lost (Fig. 2 -13B) . If the recording had been made
with this same headwheel panel, the result would be as in Fig. 2 -15B, which
is the same as was indicated in Fig. 2 -14B. In this case, however, the guide
is actually being adjusted to obtain minimum skewing of the recorded
signal. Since the recorded signal was made with a guide radius 0.5 mil
larger, the playback guide radius will be made 0.5 mil larger (0.5 mil
lighter pressure) by backing the guide away 0.5 mil from precisely con-
centric operation.
0.5us -¡
Fig. 2-15. Interchange S- distortion.
A Signal
B Signal
II
20ns-1
See Fig. 2 -15C. Note that the cancellation of timing error is not perfect,
due to the curvature caused by nonconcentric operation of the panel. As
shown by Fig. 2 -15D, the resultant "interchange S- distortion" is about 20
nanoseconds peak -to -peak at the tolerance extremes. Although the stated
guide dimensions are reasonable manufacturing tolerances (0.5 mil, or 500
microinches, is about 1/6 the thickness of this page) , manufacturers are
holding even closer tolerances. Furthermore, such errors are readily reduced
to practically zero by automatic time correction circuitry, as discussed in
Chapter 6.
Tape Vacuum Guide Too Close to Head

1. In Fig. 2 -16A, X -X -Z represents the recording dimension with a given
number of pulses in arc X -X. The playback dimension when the guide
is too close to the head is represented by Y -Y Z. Here, arc Y -Y con-
tains the same number of pulses as X-X. The pole tip now enters
early and leaves late compared to the recordini; time base.
2. Each band of lines starts early and ends late, producing leading skew
(Fig. 2 -16B), which is opposite to that of Fig. 2 -14B.
Enters Early
X-X = 16 Lines
Y -Y = 16 Lines
X -X -Z = Recording Dimension
)Z
Y -Y -Z = Playback Dimension
- - --
(A) Recording and playback dimen rions.
- - Leaves Late
Starts Early
Ends Late
\ 16 Lines
16 Lines
1
16 Lines
Leading Skew
(B) Bands of lines. (C) Eft et on vertical lines.

Fig. 2 -16. Space distortion, vacuum guide toc close to head.
3. The effect on reproduced vertical lines of space distortion caused

when the vacuum guide is too close to the head is shown in Fig. 2 -16C.
Tape Vacuum Guide Too High

Vertical alignment of the guide constitutes a mechanical adjustment
only. The correct alignment is that which allows maximum concentricity
of the tape with the pole -tip circumference. A major problem in the inter-
changeability and interspliceability of the tape is maintaining the same
stretch of tape at top and bottom of the pole -tip arc.
1. With the guide too high (Fig. 2 -17A), less stretch occurs at the top
than at the bottom; hence, there is lack of concentricity.
2. Assuming the tape was recorded with correct concentricity, the pole
tip enters late and leaves late compared to the recording time base
(Fig. 2 -17B).
3. The start and end of each band of lines is late, producing a scallop, as
shown in Fig. 2 -17C.
4. Vertical lines on a video monitor appear as shown in Fig. 2 -17D
when the guide is too high.
If a recording is made with the guide positioned as shown in Fig. 2 -17A,
it can be played back on the same head without distortion. However, if
Enters Late
Y'
(A) Guide position

Y \
Leaves Late
(B) Recording and

too high. playback dimensions.
Starts Late
Ends Late
Scallop
(C) Bands of lines. (D) Effect on vertical lines.

Fig. 2 -17. Space distortion, vacuum guide too high.
this tape is played back on a correctly adjusted head assembly, the guide
must be misadjusted vertically to eliminate scalloping in the picture. A
misadjustment such as that in Fig. 2 -17A produces excessive wear on the
bottom half of the tape.
Tape Vacuum Guide Too Low
1. When the guide is positioned too low (Fig. 2 -18A), more stretch
occurs at the top than at the bottom of the tape.
'Y- Enters Early
Leaves Early
(A) Guide position (B) Recording and

too low. playback dimensions.
Starts Early
Ends Early
Scallop
(C) Bands of lines. (DI Effect on vertical lines.

Fig. 2 -18. Space distortion, vacuum guide too low.
2. Assuming that the recording is made at the correct vertical align-

ment, the pole tip enters early and leaves early compared to the
recording time base (Fig. 2 -18B) .
3. Each band of lines starts and leaves early (Fig. 2 -18C), producing a
scallop which is opposite to that of Fig. 2 -17C.
4. Vertical lines on a video monitor appear as shown in Fig. 2 -18D

when the tape guide is adjusted too low.
Quadrature Error
In the rotary -head tv -tape system, the ideal head spacing is exactly 90
degrees plus or minus zero seconds of arc. Manufacturing problems under-
standably place considerable strain on this "perfect" situation. For example,
since there are 60 minutes in a degree and 60 seconds in a minute, there
are 90 x 60 x 60 = 324,000 seconds in 90 degrees. Table 2 -1 states that
90° of head rotation occurs in 1041.6 µs. Therefore the time required for
one second of rotation is 1041.6/324,000, or 0.00321 microsecond. If the
head is off quadrature by just 30 seconds of arc, the error is equal to
30 X 0.00321, or 0.0963 microsecond -close to 0.1 microsecond.
Correct 2 Actual
f t
Head 2 Displaced
by O Degrees
(A) Head.
Normal Normal
Starts Late -I 2
(B) Effect on lines.
(C) Appearance on monitor.
111111111111111
Fig. 2 -19. Space distortion, head -quadrature error.
If a recording is made with +30 seconds of quadrature error and played

back on a different head with -30 seconds of error;, the accumulated error
is 60 seconds (1 minute). This equals an approximately 0.2- microsecond
error in quadrature -a noticeable displacement of a complete band of lines.
In practice, quadrature error must be held under approximately ±0.02
microsecond to be invisible on vertical picture lines.
Space distortion caused by head quadrature error is shown in Fig. 2 -19.
The position of the heads on the rotating drum is shown in Fig. 2 -19A.
Notice that head 2 lags the correct position by O degrees. Assuming that
a tape is recorded with correct quadrature, the band of lines from head 2 is
displaced on playback as shown in Fig. 2 -19B. Displacement is to the right,
since the head enters the recorded information late. Fig. 2 -19C shows the
effect on a series of vertical lines viewed on a video monitor.
If a recording were made with the error shown in Fig. 2 -19A and played
back with correctly spaced leads, head 2 would be advanced in time rela-
tive to the recorded information. This deletes a portion of the duplicated
overlap interval ( Fig. 2 -9) and starts the lines of head 2 slightly early,
displacing the band of lines to the left.
The "reading" of quadrature error is complicated by certain misleading
factors, such as possible overmodulation, overcompensation of high -fre-
quency response, gap tilt, and horizontal flywheel in the monitor.
Gap 2 Azimuth Error
Fig. 2 -20. Azimuth error.
Gap 1
Other Time -Space Error Sources

Azimuth error, a manufacturing problem not subject to the control of the
user, occurs when the gap is not perpendicular to toe transverse magnetic
track made on a tape by a head with zero azimuth error. In Fig. 2 -20, gap 2
is in azimuth error. If a properly recorded tape is played back with this
error present, banding of the number 2 channel from loss of higher fre-
quencies and reduction of signal -to -noise ratio resul s. A pseudoquadrature
effect can also occur.
Axial- position error is another manufacturing defect. In Fig. 2 -21A,
head 4 is positioned to the right of its correct location. The effect of axial-
position error on a recording is shown in Fig. 2 -21B. Banding of one

channel occurs because of irregular spacing between head tracks 3 and 4.
This effect often occurs as half- banding on a single channel.
Techniques for checking possible head errors are outlined in Chapters 11
and 12. However, such techniques are full of pitfalls unless a good back-
ground in theory and practice is acquired.
4 Axial Position Error
3 ,
(A) Head location. (B) Tracks on tape.
Fig. 2 -21. Axial- position error.
2 -6. THE HEADWHEEL ASSEMBLY
Fig. 2 -22 illustrates the Ampex Mark Ten head assembly. The complete
assembly is built on a rigid base plate, and the base plate is fastened to the
top plate of the transport by one knurled and two hex-head screws. Two
integral connectors on the underside of the base plate engage mating con-
nectors integral with the top plate.
The servo -controlled, hysteresis- synchronous, air-bearing drum motor is
driven by square-wave (instead of sine -wave) ac power, generated by the
Intersync servo system, which also controls its instantaneous frequency. The
nominal frequency is 240 Hz.
The drum -motor armature assembly is supported by air bearings which
cause it to float in a nearly frictionless condition on a thin layer of air.
The air used is furnished at 55 psi (nominal) through an opening (on
the underside of the base plate) which mates with an air line from the
bearing air supply. Due to the dependence of the armature upon air under
pressure for its support, it is imperative that it never be turned or rotated
in the absence of the bearing air pressure. To do so may seriously damage
the air- bearing surfaces.
The video head drum is supported by and mounted on the armature
shaft of the drum motor. The four video heads are mounted on the head
drum in precise quadrature. The drum motor turns the video head drum
at 14,400 rpm.
Head Drum Cover for

Nuvistor Prramp
Test Points
Connector
(Under Cover Which is
Removed Upon Installation)
Vacuum Hose
Vacuum Tape Guide
Vacuum Guide Rotary

Height Knob
Transformer
ReclPlay Lever Vacuum

Guide
Vacuum Latch
Tape Guide
Solenoid
Fig. 2 -22. Ampex Mark Ten video head assembly.
The drum tachometer (described later) is located at the "upstream" end

of the drum -motor shaft. It includes a transducer which derives a con-
tinuous signal that indicates the precise instantaneous phase position of
the video head drum. This signal is fundamental to the actions of the
Intersync servo.
The rotary transformer is placed adjacent to and "downstream" from the
video head drum. Its primary windings numbers 1, 4, 2, and 3 are electri-
cally connected to video heads numbers 1, 4, 2, and 3, respectively, and
their core structure is supported by the drum -motor shaft. Hence the pri-
mary windings rotate with the video head drum. The core structure of
the associated secondary windings is supported by the base plate of the
video head assembly. The outputs of secondary windings 1, 4, 2, and 3 are
routed directly to the nuvistor preamplifiers in video channels 1, 4, 2, and
3, respectively.
See Fig. 2 -23 for the method of Ampex head identification. Note that
the head -numbering sequence is 1- 4 -2 -3. We will see later that head 4 is
gated on immediately following head 1 on playback so that the fm from
the heads is passed in sequence just as head 4 physically contacts the tape
immediately after head 1. Similarly, fm from head 2 is .gated on immediately

following fm from head 4, etc. The heads are numbered not according to
actual physical sequence, but according to gated channels. This arrange-
ment minimizes cross talk between heads. The Ampex head -numbering
sequence differs from that of RCA, but this is of no consequence in practi-
cal operations.
Channel l
H
Channel 2 -a-
Fig. 2-23. Ampex head -
identification method.
Channel 4
Channel 3
The nuvistor plug -in preamplifiers are built on one etched board
adjacent to the rotary transformer. Each is a one -stage amplifier which
functions during the system play mode and is biased off during the system
record mode.
The connector at the right end of the video head assembly (Fig. 2 -22)
carries the outputs of the individual nuvistor preamplifiers, as well as their
incoming power. This connector mates with another at the end of a cable,
which routes the preamplifier outputs to their associated preamplifiers in
the head channel housing, and also connects the power required by the
nuvistors.
The surfaces of the vacuum tape guide which face the video head drum
are curved to the circular path described by the rotating video head tips.
Vacuum is applied to a narrow cavity between these surfaces to hold and
cup the tape on this curve. A shoulder on the lower end of both curved sur-
faces stabilizes the position of the lower edge of the magnetic tape.
The vacuum tape guide is hinged to allow it to swing away from or
toward the rotating headwheel. Fig. 2 -24 shows the vacuum guide swung
back from the head on the RCA headwheel panel. When the guide is
closest to the head drum (Fig. 2-22 ) , the guide latch acts to hold it there;
when it is desired to open the guide to gain access to the video heads, the
guide latch is actuated to unlock the guide from the closed position.
On the Ampex assembly, there are two factory -set controls of the position
of the vacuum tape guide. The first is the guide- height knob; the second is
the REC /PLAY lever. When the guide is closed and the REC /PLAY lever is
set at the record position, the vacuum tape guide accurately positions the
tape for standard tip penetrations of the tape.
Tonewheel Headwheel Motor Guide Guide Height Slip Ring and Reference
(Tachometer) Motor Air Shield Release Adj Knob Brush Cover Button
Lever (Scallop)
VOA bff
Reference Vacuum Vacuum Vacuum Headwheel Control Reference

Button Hose Guide Arm Guide Trafic Head Button
Courtesy RCA
Fig. 2 -24. RCA headwheel panel assembly.
The guide solenoid advances or retracts the guide as a function of the

system mode selected. It is energized to advance the guide and establish
tip -to -tape engagement during the play and record modes, and it is de-
energized during all other modes of tape movement (such as rewind and
fast forward) Under the latter condition, a tension spring acts to retract
.
the guide and disengage the tape from the video head tips.
The RCA headwheel panel Assembly ( Fig. 2 -21) plugs into receptacles
on a recessed cast mounting on the tape -transport panel. Three captive
thumbscrews secure the headwheel panel to the :ape- transport panel and
facilitate removal and replacement. The headwheel motor rotates on an air
bearing to drive the video headwheel. A jumpe- wire connected to the
headwheel panel activates the high -pressure air system when the headwheel
panel is plugged into position on the tape-transport panel. Removal of the
headwheel panel stops the high - pressure air system from operating.
NOTE: The headwheel panel is a precision instrument that requires
special factory procedures using special tools and instruments for main-
tenance. Therefore it is strongly recommended that parts replacement be
avoided and maintenance be limited to the procedures described in the in-
struction book. If operating difficulty is encountered that cannot be cor-

rected with the cleaning and adjustment procedures presented, the head -
wheel panel should be replaced with a new or reconditioned headwheel
panel from spare -parts stock.
The headwheel is a nonmagnetic wheel having four equally spaced video
heads mounted on it. The headwheel diameter is approximately 2.0642
inches, and the video heads protrude beyond the wheel diameter by ap-
proximately 3.5 mils when the headwheel is new. The heads are mounted
with high precision to obtain a precise quadrature of the four gaps around
the wheel rim.
The video heads are 10 mils wide and consist of a coil on a trapezoidal
ferrite core, a hard spacer for maintaining a precise gap length, and metal
pole pieces of Alfecon. Resonance of the head with its load capacitance
produces a "ledge" at the carrier frequency; this normally offsets a falling
response due to head losses.

Fig. 2 -25. Ampex Mark XX video head.
The headwheel motor, headwheel, slip rings, and brushes form a single
assembly. The three -phase asynchronous motor rotates at 240 rps. It is
driven by three power amplifiers (one for each phase) in the headwheel
servo system. Fig. 2 -25 shows the physical appearance of the Ampex Mark
XX video head. This contains a special 6 -phase motor winding instead
of the conventional 3 -phase type. It is used in the Ampex AVR -1 tape
system, and the ACR -25 video cassette system (described in Chapter 13.)
The motor is surrounded by a metal shroud which serves both as a mag-
netic shield and an air duct. Cooling air is drawn in at the headwheel end
of the motor, passes over the motor housing, and i... exhausted to the head-
wheel blower through a hole in the headwheel mounting plate.
Fig. 2 -26 illustrates the RCA method of identifying the video heads.
Compare this to the Ampex method of Fig. 2 -23. Nine silver slip rings
mounted on the headwheel shaft serve to connect the video -head signals to
associated brushes and output connector pins. Each of the four heads has
separate input and return wires to isolate their respective signals and
r Rec/Play Ch 1
Changeover Relay Playback
Ch 1 _04> o t Ch 2
Rec Amp Playback
Ch 2 _o_>> o t Ch 3
Rec Amr Playback
Ch Ch 4
}
Rec Amp Playback
Ch 4 Playback
Rec Amn Brush
Channels
Returns
`
R Head
Cec urrent
Slip Rings
and
Brushes
r 9
Motor
' Headwheel
Nearest to Motor
Rotor
Head 4
Gd
Headwheel Panel
Note Changeover Relay Shown in Play Position
Fig. 2 -26. RCA head -identification method.
reduce cross talk. The rotor ground connection s made with a final con- :
nection to the ninth slip ring. Two carbon brushes associated with each
slip ring are spring loaded in pairs to equalize forces and increase reliability.
Note that the heads on the rotating wheel are numbered in sequence,
1- 2 -3 -4. However, head 1 is connected to the first pair of brushes (1 and
2 ) , while head 2 is connected to pair 5 and 6. Noie that if the RCA method
of identification was in the manner of Ampex, the heads would be num-

bered 1 -3 -2 -4 (as arranged on the slip-ring assembly) compared to the
Ampex sequence of 1- 4 -2 -3. Recordings made on either system are entirely
compatible and interchangeable; the head-numbering method is immaterial
in this respect. But the user must be familiar with the method used in his
particular system. Table 2 -2 correlates the slip-ring numbers and associated
signals.
The tonewheel and associated tonewheel head are used to provide an
indication of the speed and angular position of the headwheel. The tone -
wheel head is an electromagnet with a coaxial gap. As the tonewheel
rotates, a notch in its rim varies the reluctance of the head and produces
240 -Hz pulses in the pickup coil. The tonewheel is mounted so that the
notch is located at a precisely determined position with respect to video
head 1.
Table 2 -2. RCA Slip Rings and Signals
Slip Ring Signal

1
,nearest motor) Head 1 return

2 Head 1 input
3 Head 3 return
4
Head 3 input
5 Head 2 return
6 Head 2 input
7 Head 4 return
8 Head 4 input
9 Rotor ground
The vacuum guide assembly consists of a solid piece of nonmagnetic

material that is curved to a required inner radius. A slot at the center of
the inner radius provides clearance for protrusion into the tape of the pole
tips on the video heads. Two additional slots on each side of the center slot
provide an orifice for the vacuum side of the low- pressure air system.
Vacuum holds the video tape against the curved inner radius to form a
precise cylindrical surface for penetration by the video heads during
headwheel rotation.
The vacuum guide is mounted on a triangular -shaped arm which is
pivoted near one end of the headwheel panel. The guide end of the
vacuum guide is supported by ball -tipped assemblies that are positioned
over an adjustable wedge-type assembly when the vacuum guide is closed.
An associated thumb screw moves the wedge assembly in a perpendicular
direction with respect to the ball -tipped assemblies, thereby making the
guide arm (and video tape) perpendicular with respect to the headwheel
panel. This provides the scallop adjustment required during setup pro-
cedures.
After the vacuum guide has been locked in the general operating area
with the vacuum -guide release lever, tape -penetration pressure is provided
by the guide -arm assembly. The guide -arm assemDly is spring -loaded to
press an eccentric bearing against the edge to the vacuum guide and move
the guide and tape away from the rotating headwheel when air pressure is
removed. This is a safety feature that prevents the headwheel from dam-
aging the tape when the machine is stopped, or during other operations
which require the tape to stop while the headwheel continues to rotate.
The guide -arm assembly consists of a round shalt with an arm clamped
at one end and a bearing mounted eccentrically at :he opposite end. A flat
head at the upper end of the shaft contains an extension of the shaft near
the end of the flat head. The bearing is mounted on this extension to pro-
vide an eccentric pressure to the edge of the vacuum guide. Since the
vacuum -guide release mechanism provides a spring pressure pulling the
vacuum guide closed, the eccentric gear holds the vacuum guide open to
the required penetration depth.
Two additional bearings hold the guide-arm ass =mbly to the headwheel
panel. One is a bearing near the upper surface of :he panel and is pressed
into the panel and guide -arm shaft. A second bearing near the lower surface
of the panel is pressed to the headwheel panel and spring loaded to the
guide -arm shaft by means of a circular wedge on the shaft that is pressed
to the lower bearing by a spring and spring- retaining clamp on the shaft.
When the tape is moving, air pressure is applied to the guide air cylinder
(not part of the headwheel panel) which moves the guide -arm assembly
with an attached circular spring until the arm attached to the eccentric
shaft stops against the lead screw of the vacuum -glide positioner assembly
(not part of the headwheel panel) The position of the lead screw on the
.
vacuum -guide positioner assembly determines the amount of penetration

of tape by the heads of the headwheel.
The control -track record /playback head is mounted on the triangular -
shaped arm that holds the vacuum guide. In the record mode of operation,
this head records a control track on the tape. In the play mode, the head
picks up the control -track signal and feeds it to the servo system to provide
proper tracking between the video heads and the recorded information on
the tape.
Since the control -track record /playback head is mounted on the vacuum -
guide arm, the head is constantly in contact with the tape (even in the
stop mode). Consequently, there is a minimum of delay in reception of a
control -track pulse by the servo system.
NOTE: Further details of headwheel assemblies ai well as a study of the

associated air systems are included in Chapter 4.
EXERCISES
Q2 -1. Name the proper sequence of signal information as recorded from
"top" to "bottom" of a video tape.
Q2 -2. When the headwheel is mechanically aligned for minimum geometric
errors, is the headwheel itself physically moved?
Q2 -3. What is the head -to -tape velocity when the pole tips have a 2 -mil
projection?
Q2 -4. (A) What is the length of a tv line (1H, or 63.5 ,µs) recorded with a
head having a 2 -mil tip projection? (B) What is the length of H re-
corded with a head having a 1 -mil tip projection?
Q2 -5. If a recording is made with a pole -tip projection of 3 mils, can it be
played back on a head having only a 1 -mil tip without readjustment
to eliminate geometric errors?
Q2 -6. If a recording is made with a head having a 0.8 -mil tip projection,
will it meet SMPTE standards?
Q2 -7. What adjustment would be wrong in recording if there is a difference
in spacing for a tv line between the top and bottom of the video track?
Q2 -8. Convert 1.59 is to nanoseconds.
Q2 -9. If the raster width on a vtr picture monitor is 10 inches, what hori-
zontal displacement will occur between "bands" of picture with a play-
back error of 1 N.s?
Q2 -10. Why would the 4.2 -MHz portion of a multiburst signal break out first
at the vertical center of the image when the vacuum guide is adjusted
for lighter pressure (less tip penetration) ?
CHAPTER 3
Basic Requirements: hat the

System Must Do
A brief tabulation of what the television tape system must do can be

outlined as follows:
1. The composite video signal is presented to the four rotating heads

simultaneously in the form of an fm signal.
2. One revolution of the headwheel lays down four video tracks. Since
the tape is pulled past the headwheel at 15 in s, adjacent video tracks
are spaced approximately 5 mils apart.
3. To recover the recorded information, the video-head outputs are first
amplified, then fed to an electronic switcher which selects the signal
from the head that is in contact with the tape. This selection occurs
during horizontal- retrace intervals so that switching transients are
invisible.
4. The electronic-switcher output is then demodulated to recover the
video signal from the rf carrier. The recovered video is then processed
to "clean up" the sync and blanking intervals for distribution to the
television system.
5. For proper synchronization, the tape velocity (15 in /s) must be
locked to the headwheel rotation (240 rps) During recording,
.
stripped -off sync from the incoming video is utilized to obtain a 240 -
Hz signal (4 times a field pulse frequency of 60 Hz) , which is
amplified to obtain sufficient power to drive the headwheel. This
240 -Hz signal is also converted to a sine wave, which is recorded on
the control track of the tape.
6. Also during the record mode, a signal determined by the speed of the
headwheel is obtained and divided to 60 Hz to supply power to the
capstan motor. The speed of the capstan motor has absolute control
over the velocity of the tape, which is a nomin it 15 inches per second.
7. In the playback mode, the headwheel again rotates at 240 rps with
57
the initial reference being either the power -line frequency or local
sync. The 240 -Hz control -track signal is phase- compared to the play-
back speed of the headwheel, and any phase error slightly modifies
the 60 -Hz frequency applied to the capstan motor. Thus, the tape
velocity is continuously maintained so that the rotating heads sweep
directly across the video tracks originally laid down.
8. The audio signal is handled by conventional magnetic-recording
techniques. (The audio portion of the recording system is outlined
later in this chapter.)
3 -1. THE TAPE TRANSPORT

The tape transport conveys the magnetic tape across the recording and
pick -up heads at the correct velocity and tension. We will cover only the
basic function and orientation in this chapter. Chapter 4 describes in detail
the transport and associated equipment.
The basic transport, whether it is physically placed horizontally or
vertically, is illustrated by Fig. 3 -1. The following numbers refer to the
identification numbers given in Fig. 3 -1.
O Supply Reel
Q Upper Stabilizing Arm
0 Master Erase Head
0 Air Guides
Tonewheel and Head
Motor
Guide Shoe
s
Headwheel
Control -Track Head

Brush Assembly
Audio and Cue Erase Heads
( Guide Post
Audio and Cue

''-'' Record /Playback Heads
0 Simulplay Heads
O Take -Up Reel
Air Guide
0 Capstan and Pinch Roller
C) Lower Stabilizing Arm -

u Counting Capstan and Time Indicator
Fig. 3 -1. The basic RCA tape transport.

BASIC REQUIREMENTS: WHAT THE SYSTEM MUST DO 59
Table 3 -1. Tape -Reel Data

Reel Diameter
Feet Minutes (Inches)
1200 16 6
2400 32 8
4800 64 121/2
7200 96 14
1. The supply reel contains the tape on which a recording is to be

made or from which a recording is to be played. The length and
recording times for the various reel diameter, are given in Table 3 -1.
2. The upper (left in a horizontal tape transport) stabilizing arm pro-
vides a damping action and takes up excess tape in the event of
erratic motion, as in starting. (See Item 13.
3. The master erase head is energized (in the record mode only) to
erase any previous signals.
4. Air guides preceding and following the ma;ter erase head position
the tape for its approach to the headwheel. Air under pressure from
the headwheel blower escapes through smal pinholes in the guides
across which the tape passes. This technique is sometimes referred
to as air lubrication. (Not always used.)
5. Video and control signals are either written onto or read from the
tape by the headwheel panel assembly.
6. The audio- and cue -erase heads (energized Li the record mode only)
clear their respective tracks of any previous information and remove
the video from this portion of the tape.
7. This guide post serves to maintain tight tape contact with heads 6,
8, and 9.
8. Simulplay heads consist of an audio playback head at the top of the
tape, which functions to ensure that an audio track is being recorded,
and a control -track playback head at the bottom of the tape to ensure
the control -track amplitude is maintained. The two heads are
mounted on a single post.
9. The audio record /playback head is at the top of the post. The cue
record /playback head is at the bottom of the post. The relative
location of the sound and cue tracks is given in Fig. 2 -2.
10. This air guide changes the direction of the tape for approach to the
capstan and pinch rollers.
11. In the normal play and record modes, the pinch rollers, actuated by
a rotary solenoid through a spring -loaded linkage, clamp the tape
against the capstan, which controls the tape speed. The shaft of the
capstan motor is sized to give a tape velocity of 15 in /s at a motor
speed of 600 rpm. A hysteresis-synchronous two -phase motor with 11
inch -ounces rated running torque is employed. A flywheel is
mounted on the rear end of the capstan -motor shaft to smooth out
speed fluctuations.
12. The counting- capstan diameter and spur gears are designed to give
one count per 15 inches of tape. The counter indicates minutes and
seconds of tape, with a maximum time of 99 minutes and 59
seconds.
13. The lower (or right) stabilizing arm provides the same control
over the tape on starting the transport as does the upper arm (2) .
Both of these arms sometimes employ unidirectional air dampers

(mounted on the rear of the support panel) for proper tape damp-
ing and stabilization.
14. The takeup reel provides storage of the tape after it passes through
the machine.
The headwheel blower, mentioned under item 4, cleans and cools the
headwheel. A permanently lubricated ball-bearing motor drives a multi-
stage blower which draws the air from the headwheel panel over an oil -
wetted wire filter. This filter traps dust, lint, and iron-oxide particles from
the tape and headwheel.
In the normal play and record modes, the reel motors supply power to
hold back or take up the tape; the capstan maintains full control over tape
speed. During rewind, the upper reel motor is fully energized, and the
lower motor is energized to rotate counterclockwise (although its direction
of rotation is clockwise) to maintain tape tension and avoid spillage.
During either rewind or fast forward, the pinch rollers are lifted to
disengage the capstan.
Brakes are spring -actuated and solenoid -released for fail -safe operation.
Braking is done by an asbestos -lined band on a brass drum. The linkage is
arranged to give differential braking on approximately 240° of the drum.
Differential braking is used to keep the tape tight during stop from high-
speed wind or rewind.
The adjustment and maintenance of proper tensions on the tape trans-
port are very important to overall system stability. The techniques are out-
lined in Chapter 12.
3 -2. THE RECORDING MODE
Fig. 3 -2 is a functional block diagram of the basic system when it is in

the record mode of operation. Only the basic requirements not including
later refinements are discussed at this time. The audio section is also dis-
regarded at this time. Notice that the majority of circuits are involved in a
time ( synchronizing) function rather than in handling video. The follow-
ing numbers correspond to the like -numbered points in Fig. 3 -2.
1. The video signal is applied to a frequency modulator. The resulting
fm signal is heterodyned with a fixed oscillator.
0I
FM Head
Video
In odulator Driver
r
Sync Strip
(60 Hz)
160 -Hz Ref
® Slip Ring -Brush Assembly
Headwheel -
®Servo Units 240 Hz ssii
0
Head Motor
x4
240 Hz
240 Hz Head -Motor Head -Shaft

Ph se
Oscillator Drive Amp
® Tachometer
Comparator
240 Hz From Head Speed

® Control -
240 Hz Track
Filter
MV 240 -Hz Sine Wave Head
(Derived From Head Speed)
Capstan -
60 Hz 2$ Capstan
.4 Motor
60 Hz Drive Amp
Motor
(Derived From Head Speed)
Fig. 3 -2. Basic block diagram, recording mode.
2. The fm output from the modulator is amplified and split into four
separate channels, one for each of the rotating video heads. This is
necessary in the record mode because the video amplitude of each
head must be adjusted individually to obtain optimum results.
(NoTE: Whenever "video" is mentioned relative to the video heads,
it must be understood that the signal is in the form of a frequency-
modulated carrier.)
3. The individually adjusted fm outputs from the driver amplifier drive
all four video heads in parallel. No head-channel switching occurs
during the record mode of operation.
This completes the basic video function when recording. The remaining
functions serve in the timing of the recorded signals.
4. Sync pulses are stripped from the video signal being recorded, and
the field frequency (60 Hz) provides the -vitiating power source
for the rotating -head motor. This is termed the reference pulse.
5. The 60 -Hz pulses are multiplied to 240 Hz in frequency multiplier
A and fed into phase- comparison circuit B. The other side of this
comparator receives a 240 -Hz nominal frequency from the head -
shaft tachometer, a device for measuring shaft speed. The resultant
error signal is used to control the nominal frequency of 240 -Hz
oscillator C, which feeds the head -motor drive amplifier.
6. The head -motor drive amplifier is capable of supplying over 80
watts of 240 -Hz, three -phase power to the head motor. Nominal
power consumption is about 50 watts. The multiple -phase output is
obtained through the use of Scott-connected transformers.
7. The head motor is a three -phase 240 -Hz type.
8. The head -tachometer signal is used to generate the master -control
signal recorded on the tape and to develop the initial power refer-
ence for the capstan motor. This signal indicates the relative position
of the heads, controlled by the head motor, and the longitudinal
position of the tape, controlled by the capstan motor.
The tachometer (Fig. 3 -3) is simply a timing ring and a pickup head.
For example, the RCA "tonewheel" is a metal wheel with a notch that
breaks the magnetic path as it passes across the tonewheel pickup head,
inducing a pulse of current. The pulse is used to form a symmetrical square
wave of 240 Hz.
9. The control -track head records the signal (now a sine wave) which
is derived from the head tachometer. Since this same initial reference
from the head tachometer is also fed to the phase comparator (B of
item 5) , any tendency of the motor speed to "hunt" is compared to
the original signal timing to indicate the magnitude of error. The
corrective action of the servo controls the motor speed.
10. The capstan-motor drive amplifier receives 60 -Hz information de-
rived from the tachometer (through D and F of Fig. 3 -2) and pro-
vides amplification to the two -phase power necessary for the capstan
Timing Device on
Motor Rotating Shaft of
Headwheel Motor
Tachometer
0 Pickup Head
Fig. 3 -3. The tachometer, or timer.

Pulse
1/40
Second
Timed Edge of Square Wave

motor. Thus, the capstan is electronically coupled to the rotation of

the video heads.
11. The capstan motor is a two -phase synchronous type. It has full con-
trol over the tape speed, which is phase- lockec, to the rotating heads
in the manner described previously.
Some important relationships to remember concerning the basic record-

ing mode are as follows:
1. The video to be recorded is converted to an fin carrier, which is split

into four individual signals for feeding to the rotating heads through
a slip -ring brush assembly. The carrier amplitude for each individual
head is adjusted to obtain optimum playback of the recorded signal.
Pre -emphasis is employed to compensate for various high- frequency
losses in the head -to -tape response.
2. The speed of the rotating -head motor is made four times the vertical
frequency of the signal being recorded (60 X 4 = 240 rps) . The
vertical- scanning frequency in the incoming signal itself supplies the
initial power for this motor so that any line -to-line or field -to -field
phase drift is followed by the motor.
3. The speed of the head motor determines the absolute values of the
deflection frequencies, control -track signal, and capstan speed; there-
fore, it must be tightly controlled. A three -phase winding permits
tighter phase control with less power. The leading edge of the tach-
ometer signal bears a direct relationship to the head in contact with
the tape. This signal is continually compared with the initial refer-
ence pulse so that any tendency of the motor to hunt in speed is
corrected by an error signal.
4. The action described in the preceding paragraph assures that each
picture field occupies exactly 16 complete transverse tracks on the
tape. (Each revolution produces four tracks. This is one - fourth field,
so four revolutions equal one field, or 16 tracks.)
5. Since the speed of the head motor is the determining factor for all
frequency components, the signal from the head tachometer is con-
verted to a sine wave (nominal frequency of 240 Hz) and recorded
as a control track to supply the playback circuits with proper timing
information.
6. The head- tachometer signal is also divided down to 60 Hz to be
amplified for power to drive the capstan motor.
7. The phase -stability requirements for the capstan motor are not as
severe as for the head motor. Therefore a two -phase motor is practical.
NOTE: The foregoing discussion has omitted the audio function and the
circuitry associated with the vacuum -guide control. Audio is covered later
in this chapter, and the vacuum -guide circuitry is covered in Chapter 7.
It should be understood that the method of attack in system functions is,

in some instances, quite different between Ampex and RCA practice.
Except where noted, the discussions in this section are general in nature
rather than specific. The first consideration is to become introduced to
fundamental working requirements.
For a more complete diagram of the basic function of modern servos in
the record mode, substitute the drawing of Fig. 3 -4 for the dashed -in area
of Fig. 3 -2.
Stripped Sync
Field -Pulse
Head -Motor To
From
Add Control Hea 'Motor
Incoming Video Gen Error
(or Local Sync
Circuits Drivers
Signal
for Local
Phase Gain
Recording)
60 pps Velocity Gain
240 Hz
Phase Velocity
Comp rator Detector
Lock Bus
240 Hz 240 HzFrom

Tachometer
Frame Pulse Con rol - To

From Ref Gen Trac Gen ,4
Capstan -Motor
30 pps 160PPsl
240 + 30 pps Drivers
240 -Hz Sine Wave + Franse Pulse To

Control -Track
Head
Fig. 3 -4. Basic servo units (record mode).
Since the speed of the head motor (nominally 240 rps) determines
absolute values of deflection frequencies ( video scanning) , control -track
signal, and capstan speed, this reference (pulse from head tachometer) is
made as "tight" as possible relative to the recorded video signal.
Follow Fig. 3 -4. Pulses are formed from the stripped -off sync of the
incoming signal (being recorded) at the nominal field rate of 60 Hz.
These pulses feed one side of a phase comparator. The other side of the
phase comparator receives the nominal 240 Hz from the head tachometer.
This normally is a symmetrical waveform from a multivibrator controlled
by head speed.
A velocity detector also receives the nominal 240 Hz from the head
tachometer. The velocity detector compares this signal to a fixed reference
(not dependent on any external signals) which has a period equal to that
of the nominal 240 Hz.
Note that each leading edge of a head-speed pulse is equivalent to
one revolution of the head (360 °), or an average of 65.6 picture lines,
one-quarter of a field, or 4166 µs. But the square wave (as shown directly
BASIC REQUIREMENTS: WHAT THE SYSTEM MUST no 65
from the timing ring, Fig. 3 -3) that results gives a reference only for each
180° of rotation, or 33 lines.
The added outputs of the phase and velocity circuits of the servo feed a
head -motor control circuit which also receives the nominal 240 Hz from
the head tachometer. Head -motor control might be a voltage-controlled
oscillator or a headwheel "frequency" modulator.
The head -tachometer signal, which indicates the instantaneous speed and
phase of the head relative to the recorded signal, is fed to the control -track
recording head by way of a generator which mixes a 30 -Hz frame pulse.
This pulse is called the edit pulse, or the control -track frame pulse.
A 60 -Hz sine wave is derived from the head-tachc meter signal and used
to lock the capstan to the headwheel speed.
Now go to Fig. 3 -5. Waveforms 1 and 3 in Fig. 3 -5A show the lineup
of the leading edge of the headwheel -speed pulse with the fixed reference.
This is the function of the velocity loop in the headwheel servo. Waveform
2 is the resultant output of the head -speed multivibrator.
When the headwheel is slow (as in start -up) , the pulses from the tach-
ometer ( waveform 5, Fig. 3 -5B) will be farther aparr than the pulses of the
240-Hz reference (waveform 6) An error voltage which depends on the
.
difference between the two pulses is generated and Nerves to speed up the
headwheel. When the two pulse trains reach identi:al periods, zero error
voltage results, and the headwheel speed is maintained. This is done by a
pulse -sampling technique (Chapter 7) .
The velocity -lock condition (waveforms 7 and 8, Fig. 3 -5C) means that
an interval of about 33 picture lines is locked to a fixed reference within
the velocity loop. You can see that this is a "long-time- constant" afc loop.
Another way of looking at this is that the overall "horizontal rate" is an
integration of the horizontal rate of 33 picture line;.
When this lock is obtained, the phase loop is enabled to compare the
60 -Hz (field) reference from the signal being recorded to the horizontal
envelope of about 33 lines. This 60 -Hz sampling will then be on every
fourth pulse of the 240 -Hz signal. See waveforms 9 and 10 of Fig. 3 -5C.
This is the signal normally observed on the monitoring cro in the head -
wheel servo position.
3 -3. THE PLAYBACK MODE
More circuitry is involved in the playback mode t1 an in the record mode.

There are two reasons for this. (1) Electronic switch! ng is required to select
only the head that is in contact with the tape. This prevents spurious signal
and noise outputs from being fed to the demodulator from the high -speed
heads not in direct tape contact. (2) An additional servo control is neces-
sary for the capstan motor. The speed of the tape must be controlled by a
continuous comparison of the control track laid down during recording
(240 Hz) with the 240 -Hz signal derived from the head-shaft tachometer.
360°
65.6H
I 0.25V I
Headwheel 1240 pps l
l 4166ps f
Leading Edge From

Head Tachometer
t i I
Multivibrator Controlled From

@I?- Head Speed (VCO)
(z) Fixed -Reference MV
Velocity Reference I
60PPs Field Pulse From

> Signal Being Recorded
1 Field I
16,667 ys
1440°
NA. - Velocity Servo Sampling

Q - Phase Servo Sampling
(A) System waveforms.
Headwheel at Start
Fixed ReferencelVelocityl
I--- 416616 -{ I
(B) Pulses at start -up.
Headwheel Velocity Locked
Headwheel Phase Locked
60 Hz
(C) Phase lock.

Fig. 3 -5. Velocity and phase lock of headwheel (record mode).
System Outputs
0 111
P ocessi g
Switcher FM Demod
Amp
240 Hz
(Gate)
Separated Sync
I
480 Hz From Blanking Switcher Switch Pulse)

Heads (Under Blanking on Leading Edge)
60 Hz Fr )m Places Switch
240 Hz
Head -Motor
Motor Local Sync Blanking Pulse Under
34, Drive Amp
Generator or Switcher Horiz
Power L ne Blanking
U
Tach
240 Hz From Head Speed 480 Hz
/Control -
Tread
\Head
-240 Hz
Phase Shift
(Tracking) 240 Hz
Osc
x4
=
240 x 2
480 Hz
Sauare Wave
m
Capstan
Servo 240 Hz
-Tape 240 Hz
Phase
O 16OHz Comp
Capstan 60 Hz
Cap tan
Motor o 0
Motor
24s Drive Amp
Fig. 3 -6. Basic block diagram (playback mode).
This ensures that the video heads "track," or pass directly over the recorded
video tracks on the tape. Failure of such control causer the relative positions
of the heads along the length of the tape to drift. Thus, a head could bridge
two adjacent recorded tracks rather than remaining over a single track.
A block diagram of the basic playback system is shown in Fig. 3 -6. In
the following explanation of the playback system, the numbers refer to
those used to identify the various blocks in Fig. 3 -6
1.The relatively low -level (about 2 to 5 millivolts) output from the
head slip -ring brush assembly is amplified in individual preamplifiers
before being fed to the switcher unit. For simplicity, the preamps
are not shown in Fig. 3 -6. The switcher combines the individual
head inputs into a sequential and continuous single output. A 240-
Hz signal from the head tachometer presets the gates. A 480 -Hz
signal, derived from the same source but with the leading edge
placed under horizontal blanking, is used to trigger the actual switch.
This technique is further explained in the discussion for item 14.
2. The fm demodulator extracts the useful video signal from the fre-
quency- modulated carrier.
3. The processing amplifier cleans up the blanking and sync intervals

of the demodulated video.
4. Either the local sync generator (color) or the 60 -Hz power line
(optional for monochrome) is used as the initial reference for the
head-motor power.
Caution: This does not mean that the reproduced signal from the
tape system can be treated as a local signal for lap -dissolves or local
superimposition. Neither horizontal- nor vertical-sync pulses from
the tape are necessarily coincident in time with the external pulses
from the local sync generator. Unit 4 is used only for an initial
yardstick to supply a stable comparison.
NOTE: All modern recording systems do permit local lap- dissolves, supers,
and special effects. At this time, however, we are studying only the most
basic requirements of the magnetic video recording system. The details
of refinements which allow this type of operation are in Chapter 7. The
basics are outlined in Section 3 -4.
5. The head motor is controlled by the same servo system that is used
in the record mode, except that the initial reference signal is the
local sync generator (or 60 -Hz power line) .
6. The head -motor drive amplifier serves the same function as for
recording.
7. The head motor functions the same as for recording.
8. The signal from the head tachometer is routed to units 1, 5, 10, and
13, as shown in Fig. 3 -6.
9. The control -track head reproduces the recorded control track.
10. The capstan servo compares the control -track signal (which is indi-
cating the relative position of the heads during recording) with the
head-motor tachometer signal (which indicates the playback speed
and phase) From this comparison, an error signal is developed
.
which slightly modifies the frequency of the 60 -Hz oscillator used

as the initial power source to drive the capstan motor. Note that a
control ( termed the tracking control) is provided to shift the phase
of the control -track signal. If, for example, head 1 is reproducing
vertical sync, tracking can be shifted so that head 2, 3, or 4 repro-
duces vertical sync. This allows matching of heads to obtain optimum
reproduction from a tape recorded by a different head.
11. The capstan-motor drive amplifier serves the same function as for
recording.
12. The capstan motor serves the same function as for recording.
13. A 480 -Hz square wave is formed from the 240 -Hz tachometer signal.
14. The leading edge of the 480 -Hz square -wave switch pulse is made to
occur during the horizontal-blanking interval. This prevents switch-
ing of channels during an active line interval; such switching results
in transients in the picture.
Fig. 3 -7 shows the basic principles of the switching function.

Switcher -channel inputs are gated by two 90 °- related, 240 -Hz
signals. Two 480 -Hz, 180 ° -related pulses chop the signal at a 960 -Hz
rate to place the head outputs in the proper sequence in a single
channel. (Details of this operation are given in Chapter 5) . NOTE:
The techniques used by Ampex and RCA differ, but the principle is
the same.
Time
0°
1
90°
2
180°
1 3
270°
I 4
360°
o-
1 po 240 Hz
Phase
O
1-
Gate Spl it
Gated by 3 Boo 90° Delay

240 Hz at 90° at 240 Hz
Gate 2 4 - in Time
29D° 11043 ys)
Phase
Split 90°
L
J
Switch
U On Off On Off o
4 270°
480 Hz
(B
Switch 1
1 Off On Off On
Switcher Output 1 I 2 I 3 I 4
-I I-1/960 Sec
Fig. 3 -7. Basic playback- switching functions.
3 -4. SYNCHRONOUS PLAYBACK
First of all in playback requirements, remember the information as it is

timed in recording: the headwheel must rotate one-half revolution (33 TV
lines, or 1/8 field) before a timing reference is encountered to control
headwheel velocity. When the tape is played back in the same manner as
described above, there is no lock with station sync.
The tape sync timing in playback varies considerably relative to fixed
station sync. The afc loop in the velocity servo has a long time constant
which only follows the average drift across approximately 33 lines of
picture information. This, in modern terminology, is a rather "loose"
servo. In this internal mode of playback operation, there is no line -to -line
feedback information from the tape signal itself to control the playback
headwheel velocity.
If you have access to a recording system, you can demonstrate this to

yourself by locking the picture monitor to local sync, while displaying the
tape playback signal on the same monitor. This is why local sync generators
cannot be genlocked to a tape playback in the internal mode of operation.
The jitter relative to local sync exceeds genlock requirements. It is also
why, when it is necessary to handle the tape playback signal for lap-
dissolves or special effects, the "reverse genlock" principle must be used.
The tape playback system itself must be locked and phased to the local
sync generator.
Typical side -to -side (horizontal) jitter and drift of internal -mode play-
back of video tape is around 0.1 to 0.2 p.s (100 to 200 ns), relative to
station sync. This is the stability of the servos alone, to which you must add
any geometric errors resulting from imperfectly aligned systems and normal
tolerance of components. These are velocity errors resulting in line -to -line
errors in the reproduced image. Obviously, in the case of geometric error,
there are line -to -line errors in the signal sync time -space relationship. Now
group all of these into one term: time -base error.
Let us see what the playback requirements are, to get even the very
minimum stability of time base necessary to treat the tape signal as a local
signal. This means "genlocking in reverse," or making the signal from the
tape playback system coincident in time and stability with the local sync
generator.
See Fig. 3 -8. The first requirement is that the signal from the tape system
be phased and then locked to the vertical rate of the local sync generator.
Tape vertical alignment (tva) occurs when the tape -signal frame rate
equals the reference vertical frame rate, which is the local sync rate. To do
this, feedback information from the output signal of the tape system must
be used to control the headwheel and capstan. This feedback information
must contain a comparison between the signal coming from the tape and
the local sync signal.
The added circuitry over that of Fig. 3 -4 is enclosed in dash lines in
Fig. 3 -8. In the headwheel servo, the demodulated tape sync is compared
to the local reference vertical sync. So, the signal input to the phase com-
parator now consists of an output from a variable delay circuit, which in
turn is dependent upon coincidence between local and tape vertical -rate
pulses.
In the capstan servo, the control -track signal (240 Hz) is divided (by 8)
to the frame rate of 30 Hz, which is reset by the 30 -Hz tape -frame pulse
for coincidence. The tape-frame pulse is generated from the vertical -sync
interval of the tape being played. A reference frame pulse is generated from
both horizontal- and vertical -rate pulses of the local sync generator. The
horizontal rate must be used to sense the field used for frame reference; in
the RCA system, this is field 1, in which a full line precedes the vertical
interval. Field 2 has I/2 line preceding the vertical interval; this is the refer-
ence field for the Ampex system.
r
Local Error Signal
Add
Sync
-i-
Field Pulse Head Motor
Ph se HVelocity Control
Gn Detector
Comparator Circuits
60 Hz
240 Hz
Variable
De ay
240 Hz Fron Head Tachometer
60 Hz
1
TVA Contro -Track
Control -Track Tape Frame
Phase Puls Gen
Play (Head) Pulse
Comparator 240 Hz
-r1 Coinc
Vert
dente
Detector
Contro -Track
Phase
- +8
30 Hz
.F_
30 Hz
Reset Pulse
Lock Sense Coincidence

¡racking
Sense
Field
Ref Frame 240 Hz Osc
Pulse Comp rator
Pulse NCO)
Gen Error
Volts
fi iv tH 1
Demodulated
Tape Sync
Input
Local
Composite
Sync
To
Capstan -Motor 4
Drive Amps
-
60 Hz s4
Counter
Added Circuits (Headwheel Added Circuits (Capstan)
Fig. 3 -8. Basic servo units (vertical alignment) for playback.
A voltage -controlled oscillator (vco) running at a nominal frequency of

240 Hz is varied in frequency in accordance with the amount of error
generated in the comparator. Since the tape -frame pulses have a definite
time and phase relationship with the head tachometer during recording,
the capstan is "slipped" across the heads until the head playing back verti-
cal sync is "tracked," in phase with local vertical -rate pulses.
This type of coarse vertical framing permits roll -free switching from
local to tape sources, but there is still no coincidence with local horizontal-
rate timing. Thus, additional tape horizontal alignment (tha) circuitry is
required for feedback and controlling information. This will be discussed
in Chapter 7. Assume for the moment that we have a tape playback signal
which is "genlocked" by the local sync generator.
Now what have we done thus far in time -base stabilization? We have
"tightened up" the playback servos so that we should be able to use the
tape signal as we do any other conventional studio gear, such as live

cameras, and film cameras. We can mix or lap- dissolve the tape- system
output with any other local signal source.
In reality, we may or may not have achieved this, depending on the
degree of geometric error remaining. This is why modern tape systems
incorporate video time -base correcting devices which are employed in con-
junction with tva and tha. Remember that tva and tha are servo functions
only; they do not involve sensing errors of the headwheel or drum which
result from imperfect alignment with the signal recorded on the tape (the
geometric errors of Chapter 2) The correcting devices are Time Error
.
Compensation (Amtec) for Ampex and Automatic Time Correction

(ATC) for RCA. Either of them reduces residual errors to less than 0.03
µs, or 30 ns (when combined errors are within 1 )us initially) .
It is obvious that a 30 -ns jitter, when the picture is locked to the local
sync generator horizontally and vertically, is not observable. This amount
of jitter represents about one part in 2000 at the tv line rate of 15,750 Hz.
But what about the color time base? The color -sync burst contains a fre-
quency of approximately 3.58 MHz. This burst is the phase reference for
all the chroma information carried in the sidebands of the chroma sub -
carrier of the video signal. At 3.58 MHz, one cycle represents approximately
1
3.58 (10 °) = 0.28 µs = 280 ns
Then
280
0.8 ns per degree
360
Now, if you have a color -tape signal that exhibits as much as 30 -ns jitter,
there are 30/0.8 = 37.5° of burst jitter. Quite simply, this means the color
phase will jitter over 37° to 38° of the range, which is intolerable. So for
color, additional video time -base correction units are required. These units
are Color Time Error Correction (Colortec) for Ampex and Color Auto-
matic Time Correction (Color ATC) for RCA.
In practice, the monochrome error correction must be made first, then the
color error correction. Timing errors can now be reduced by this combina-
tion to about 2 ns, or approximately 2.5° of jitter at 3.58 MHz. Since the
jitter at the color -subcarrier time base must be about 5° to be noticed, the
result is quite satisfactory.
3 -5. THE AUDIO SYSTEM

In the previous drawings of the tape transport, it was noted that the
audio heads are spaced in distance (and time) from the video heads. This
distance is 9.25 inches, and since each picture frame occupies one-half inch
along the tape, the corresponding sound for a given picture frame leads
the picture by 9.25 X 2, or 18.5, frames. Perfect lip sync is maintained

because the same audio head is used for both recording and playback. The
only problem which arises is the editing ( splicing of the tape between ;1
rapidly spoken words (or continuous music) so that proper picture con-
tinuity can be maintained.
Since the tape speed is 15 in /s, the spacing in time between the sound
and the corresponding picture frame is equal to (9.25/15), or 0.61 second.
It is obvious, therefore, that for most editing, the first frame pulse following
a spoken word may be selected for the splice (described in Chapter 10) .
Conventional audio magnetic-tape methods are used for the sound track.
A plot of the amount of remanence as a function of the strength of the
signal field is given in Fig. 3 -9. This type of transfer curve has no linear
portion that would allow an undistorted signal output. Therefore a high -
frequency bias (between 85 and 100 kHz) is added to the audio current,
as shown.
Ideal
(Linear)
Slope
Signal Field + Zero Axis - --A".B

Actual Slope -L- -L D
Undistorted Output
AF
\ AF
Bias (Supersonic)
Fig. 3 -9. Effect of supersonic bia ;.
The audio component on the plus side of the zero magnetization axis
results in a distorted waveform in which loop A is larger than loop B. The
audio component on the minus side results in a distorted wave in which
loop C is smaller than loop D. The algebraic sum of A + C equals the
algebraic sum of B + D when the loops are combined on the axis of opera-
tion, resulting in an undistorted output.
The axis of easy magnetization of the iron -oxide particles on video tape
is aligned transversely across the direction of travel to favor video response.
This places a slight limitation on audio bandwidth and signal -to -noise
ratio, as compared to conventional 15 -in /s audio tape recording. It should
also be noted that the audio track on the video tape is only about 70 mils
(0.07 inch) wide as compared to the conventional quarter -inch audio tape.
Review Fig. 3 -1 for the audio head location. Fig. 3 -10 illustrates the
details of the post assemblies and the sequence of the upper and lower tape -
section heads. Note the dual function of each post. Note also that a cue -
track simulplay head is not provided. This position (bottom of third post)
is occupied by the control -track simulplay head. The control -track record
head is on the vacuum guide (Figs. 2 -24 and 3-1.)
Fig. 3 -11 shows the basic function of the rec /play changeover relay for
audio in any tape system. The relay contacts in this drawing are shown in
the play position. In this position, the contacts act as follows:
Tape Motion --a-
Audio Record/ Audio Simulplay
Audio Erase Play Head Head
1
F ) i
70 Mils
Erase Post Guide Record /Play Sim ulplay

Post Post Post
20 Mils
I r r I 1
1
Cue Erase Cue Record/ Control -Track
Play Head Simulplay Head
Fig. 3 -10. Audio post and head arrangement.
Audio
Simulplay
Head
Thb L0
6
7
Audio
Preamp
- Audio
Playback
Out
. o
Audio
Record /Play
Head
10
11
o
4 } 12 Bias
HF Bias
Audio
Equalizer Trap Control
Play
Record
F
Record Audio In
Fig. 3 -11. Audio record /play relay (playback position).
1. The audio record /play head is connected to the preamplifier input.

This is done through contacts 3 and 9.
2. The circuit of the simulplay head is opened by contacts 1 and 5.
3. The output of the high- frequency equalization circuit is grounded
(through contacts 2 and 7) to prevent cross talk.
When in the record mode, the relay contacts accomplish the following:
1. Contact 4 connects to contact 12 to feed audio and bias into the
record /play head. The other side of the head returns to ground
through contacts 3 and 10.
2. The audio simulplay head is connected (throt.gh contacts 1 and 6)
to the input of the playback preamplifier to permit monitoring. The
simulplay head reads the audio (delayed) as it is laid down on the
track by the record head.
NOTE: The high- frequency equalizer in the recording circuit is switched
when the speed is changed (15 to 71/2 in /s and vire versa) , to provide
proper equalization for the selected speed.
Maintenance, measurements, and adjustments of the audio system are

covered in Chapter 12.
EXERCISES
Q3 -1. What is your visualization of "video" relative to the video heads?

Q3 -2. In the record mode, what signal supplies the initial timing informa-
tion for the headwheel motor?
Q3 -3. In the record mode, what signal supplies the initial timing information
for the capstan motor?
Q3 -4. In the record mode, what are the horizontal and vertical time bases of
lockup?
Q3 -5. In a nonsynchronous playback mode, what are the horizontal and
vertical time bases of lockup?
Q3 -6. In the framing (switchlock) playback mode, what are the horizontal
and vertical time bases of lockup?
Q3 -7. In the synchronous (genlocked) playback mode, what are the hori-
zontal and vertical time bases of lockup?
Q3 -8. What is the time base of (A) monochrome ( coarse) time -error cor-
rection and (B) color error correction?
Q3 -9. In the basic (nonsynchronous) playback mod «, what signal supplies
the basic timing information for the capstan motor?
Q3 -10. What is the sequence of operation of the phas«. and velocity loops of
a headwheel (drum) servo in either the record or play mode?
CHAPTER 4
The Tape Transport, Air Systems,

and Control Circuitry
The RCA TR -70 tape transport with head cover removed is illustrated
in Fig. 4 -1. The Ampex VR -2000 tape transport was shown in Fig. 1 -12.
4 -1. FUNCTIONAL DESCRIPTION OF TAPE TRANSPORT

The tape- transport system is contained on the tape -transport panel,
which is waist high and sloped back at a 45° angle to facilitate tape load-
ing. The structural support for the tape -transport panel consists of a thick
cast aluminum plate which is ground smooth on both sides. The front
surface of the plate provides a reference plane from which the mounting
dimensions of the various mechanical assemblies are calculated.
The following items are mounted on the tape -transport panel:
1. Drive and brake assemblies for the supply and takeup reels.
2. Capstan and pinch-roller assemblies.
3. Counting roller and counter.
4. Take -up and supply tension arms.
5. Vacuum -guide positioner and guide solenoid.
6. Selective erase head and tape lifter. The term "selective" is used
because sometimes only portions of the tape are to be erased; for
example, in editing procedures (Chapter 10 ) , video is sometimes
erased without erasing the control track.
7. Audio heads and audio preamplifier module (including the record/
play changeover relay) .
S. Mounting facilities and connections for the headwheel panel.

9. Air guide.
10. Head -hours meter.
Tape motion is controlled by three independent motors which drive
the supply reel, the capstan, and the take -up reel. During record or play -
76
THE TAPE TRANSPORT, AIR SYSTEMS, AND CONTROL CIRCUITRY 77
Picture Waveform -
Audio Monitor Standards Monitor
Selector Selector Audio VU Selector Picture T ode Waveform Selector
Switches Switches Speaker Meter Switches Monitor Multimeter Inc cator s Monitor Switches
Record
Control
Panel
Supply
'- " lr
r
Play
Control
Panel
Take -Up
Reel Reel
Air Selective Headwheel Audio/Cue Audio/Cue Audio and Capstan Pinch Counter Counter
Guide Erase Assembly Erase Red Play Control-Track Roller Roller
Head Head Head Simulplay Head
Fig. 4 -1. RCA TR -70 tape transpoirt.
back, the tape speed is controlled by the capstan; the supply and take -up
reels merely provide constant torque. In the wind mode, the tape is pulled
by the take -up reel, and the supply reel provides tension by pulling in the
opposite direction. The reel brakes are applied automatically whenever the
machine is placed in the stop mode. A foot -switch, located at the bottom
center of the front of the machine, permits the brakes to be released so that
the reels may be turned manually during tape threading.
The tension arms, located below each reel, take uá excess tape if speed
variations occur due to changes in operating modes,. Tape -break switches,
operated by either the take -up or supply tension arm, automatically stop
the machine if the tape breaks or if the supply reel runs out of tape. A
stationary, air-lubricated tape guide below the tension arm of the supply
reel positions the tape for its approach to the headwheel. The air guide
consists of a post that is perforated with tiny air holes and provided with
flanges at the edges. Air is forced through the holes to lift the tape slightly
off the surface. The flanges provide the guiding as Lion. (In some more
recent systems, the air holes have been eliminated, and rotating guides are
used instead.)
Tape from the supply reel proceeds under the tension arm, over the air
guide, past the tape lifter to a two- inch -wide selective erase head, and then
to the headwheel, which rotates at right angles to the tape motion. Below
the headwheel, a vacuum guide establishes suction which curves the tape
transversely to fit the wheel and holds the tape tightly so that the video
heads contact the tape. The guide position, which determines the head -to-
tape pressure, is determined by a lead screw driven by the vacuum -guide
positioner motor behind the tape -transport panel. The motor is controlled
by the guide servo system (described in Chapter 7.)
After leaving the headwheel, the tape passes under the control -track
head, which records or plays back a 240 -Hz sine wave mixed with narrow
30 -Hz frame pulses. The 240 -Hz signal is used during playback to establish
the timing of the capstan servo system so that the video heads scan along
the recorded tracks. The frame pulses permit locating the correct points at
which to cut the tape during tape splicing, and they serve other uses in the
servo systems.
Next, the tape passes under three supports in sequence; each support
contains two heads concerned with tracks at opposite edges of the tape. The
first support contains the audio /cue erase heads, which remove video signals
from the appropriate areas to provide space for recording on the audio and
cue tracks. The second support holds the cue and audio record /playback
heads. The third support holds the simultaneous playback heads which
permit monitoring the audio and control tracks during recording. Because
of the distance between the headwheel and the audio and cue heads, the
two sound tracks are recorded 9.25 inches ahead of the corresponding
video tracks. Located between the audio /cue record /playback head post and
the simultaneous -playback head post is the tape-edge guide post which
produces the required tape "wrap" around each head.
The tape then proceeds to the capstan and pinch roller. The pinch roller,
which is actuated by a rotary solenoid during recording and playback, pro-
vides friction which enables the capstan to pull the tape. The capstan motor
is driven synchronously by the capstan servo system so that the tape motion
is related to the headwheel motion. The tape speed is either 71/2 or 15
inches per second as determined by a speed selector switch. The capstan
motor drives a belt connected to the capstan flywheel and shaft.
After leaving the capstan, the tape passes under the counting roller and
the take -up tension arm to the take -up reel. The counting roller is geared
to a four-digit resettable counter which indicates elapsed time during
recording and playback and permits locating desired portions of the tape
during rewind. The counter indicates minutes and seconds and has a capa-
city of 99 minutes and 59 seconds, at a tape speed of 15 in /s.
4 -2. AIR SYSTEMS
It is probably apparent at this point in our studies that air systems play
a vital role in the tape- recording system. The description of a specific system
(RCA TR -70) will serve to give the student a practical view of the quad -
ruplex -tape air system.
The machine performs a number of functions with the following four

basic air systems:
1. Low- pressure air system for air and vacuum guides and tape lifter.
2. High -pressure air system for air bearing on headwheel and vacuum -
guide air cylinder.
3. Cooling blower for heat removal.
4. Headwheel blower for cooling the headwheel motor.
The high- and low- pressure air pumps and cooling blower are located
on the lower section of the machine, and are accessible from the front or
rear. Other parts are mounted on the front and rear of the tape -transport
panel and behind the play control panel.
Low -Pressure Air and Vacuum System
The low- pressure air and vacuum system (Fig. 4 -2) provides air for the
air -lubricated tape guide, vacuum for the tape lifter on the tape -transport
panel, and vacuum for the vacuum guide on the headwheel panel. Both the
air and vacuum functions are performed by a single pump of the oil -free,
rotary -vane type. The pump motor directly drives a slotted rotor which is
mounted eccentrically in a cylindrical air chamber. As the rotor rotates,
centrifugal force causes four carbon vanes on the rotor slots to press against
the cylinder wall. This action creates a vacuum at the pump inlet and drives
compressed air from the outlet. The pump is provided with a filter /muffler
and relief valve at the inlet (vacuum) , and a muffler, filter, and relief
valve at the outlet (air) .
Vacuum System -Air drawn in through the slots in the vacuum guide
on the headwheel panel flows through a short hose on the headwheel panel
to an outlet below the panel mounting plate. A long hose runs from this
outlet in the back of the machine, through a T connector on the rear shelf
frame, then to the vacuum filter/muffler behind the play control panel.
Vacuum passes through the filter/muffler and through a four -outlet con-
nector on its intake side. The vacuum gauge monitors the vacuum level at
this point, and a vacuum relief valve is connected to another outlet of the
connector to permit adjustment of the vacuum level. The final outlet of
the four-outlet connector is attached to a rubber hose that runs to the
lower section of the machine. Here, the vacuum l..ne is filtered prior to
being connected to the intake side of the low- pressure air motor or vacuum
pump. Thus, the vacuum guide obtains vacuum immediately when the low -
pressure air pump is energized, and it holds that vacuum in all operating
modes except when the motor is de- energized.
Exhaust Air System -Air supplied through the tir guide on the tape -
transport panel is fed from a hose attached to an outlet on the lower side
of the panel. The hose extends to a T connector on the rear shelf frame,
then to a low- pressure air filter behind the play control panel. After filter-
ing, the air is supplied through a four -outlet connector and to solenoid
r--
i
r---------
I
- - -1 0.
E
t,
r-----t
L__
T--I.
am* ION
Courtesy RCA
Fig. 4 -2. Low -pressure air and vacuum system.
valve 1L1 (Figs. 4 -2 and 4 -3) The two remaining outlets of the four -
.
outlet connector are connected to a low- pressure ail gauge and an air relief
valve, respectively, so that air pressure may be monitored and adusted as
required.
Air solenoid 1L1 is energized in any running mode of the machine ( all
modes except stop, standby, or setup) and permits air to be supplied to the
air guide. When the machine is switched from a running mode to one
which stops the tape reels, a time -delay circuit in the control module de-
energizes the solenoid after a delay of approximately four seconds. This
Low -Pressure Low -Pressure High -Pressure Regulator Vacuum
Air Solenoid (1L11 L Air Gauge Adjustment Screw Gauge
Low - Pressure Air -Relief High - Pressure high- Pressure

Valve Adjustment Solenoid (1L41 Air Gauge
Fig. 4 -3. Air system (front view) behind plag control panel.
delay permits the air to continue to flow through the air guide until the
tape reels have come to a full stop, thereby providing air lubrication at the
air guide while the tape is still moving.
An air muffler is attached to one side of the solenoid valve (1L1) to
reduce noise. The third outlet of the solenoid valve provides a connection
for the hose that runs from this outlet to the air exhaust filter on the air -
supply side of the low-pressure air or vacuum pump. With the motor of
the low- pressure air and vacuum pump energized in all running modes of
the machine, lubricating air is supplied to the air guide at all times that
the tape is in motion.
Tape Lifter -The tape lifter on the tape -transport panel is activated by
both air pressure and vacuum, which are supplied through solenoid valve
1L6 on the rear shelf frame. Each of the two T connectors mentioned pre-
viously in the descriptions of the vacuum and air flow have one outlet
connected to solenoid valve 1L6. Only during the wind mode of operation
is the solenoid valve set to permit exhaust air to flow through the solenoid
valve and into the tape lifter. This causes a sapphire rod to extend length-
wise from the tape lifter on which the tape rides and thereby prevent the
tape from passing across the selective erase head. When the machine is
switched to a mode other than wind, the vacuum line is connected through
the solenoid valve, and the air line is cut off from the tape lifter. The
vacuum presents an assist force to the tape lifter and causes the sapphire
rod to return immediately to its inserted position, thus causing tape erasure
to commence with minimum delay upon resumption of tape movement.
High- Pressure Air System

The high- pressure air system (Fig. 4 -4) provides compressed air to
function as an air bearing for the headwheel motor and to operate the
vacuum -guide air cylinder. The system contains a compressor, air tank,
tank pressure regulator valve, tank pressure gauge, air -sensitive pressure
switch, high -pressure regulator and gauge, two selective air solenoid valves,
air switch, output air filter, headwheel air valve, vacuum -guide cylinder, and
two drain solenoid valves.
The compressor, air tank, tank pressure regulator valve, tank pressure
gauge, air -sensitive pressure switch, and two drain solenoid valves are all
mounted on the lower shelf of the machine. The air switch, high -pressure
air regulator, high -pressure gauge, and one solenoid valve are mounted
behind the play control panel. The output air filter, headwheel air valve,
and vacuum -guide air cylinder are mounted on the tape-transport panel,
near the headwheel panel. The air solenoid valve for the vacuum -guide air
cylinder is mounted on a rear shelf of the machine frame.
Air -Tank Control -With the compressor operating, air is supplied to
the tank through a rubber hose and the tank pressure regulating valve.
Tank air pressure is controlled with the tank pressure regulator valve and
monitored with the tank pressure gauge. A T connector is attached to an
outlet at the lower part of the air tank. One outlet of the T connector is
attached to a rubber hose that extends to the air -sensitive pressure switch
(1S4) . If the tank pressure exceeds the pressure limitation set on the
switch, the switch opens its electrical circuit and prevents current from
being supplied to the compressor, thus providing an on -off control for the
compressor.
A rubber hose extends from the final outlet of the T connector to a
drain solenoid valve (1L2) . The solenoid valve remains closed while the
L_
L_
Courtesy RCA
Fig. 4-4. High -pressure air system.
high -pressure system is in operation; however, when the system stops, the
solenoid valve opens and permits any accumulated water to drain from the
air tank to a pan below the valve. Under normal conditions, the pan accum-
ulates a small quantity of water which is evaporated and does not necessi-
tate draining the pan.
Supply for Headwheel Air Bearing-High -pressure air is supplied from
the air tank through a rubber hose to a T connector behind the play control
panel. The T connector is attached to the high -pressure regulator, which
may be adjusted to the desired level. Two outlets are provided on the high -
pressure regulator. One outlet is connected to air switch 1S2. If the air -
bearing supply to the headwheel fails, this switch places the machine in
the stop mode to prevent the headwheel motor from being energized, and
the red AIR PRESSURE indicator above the tape -transport panel is illumi-
nated. The high -pressure gauge obtains its input on the opposite end of
the air switch.
The second outlet of the high -pressure regulator is connected to air
solenoid valve 1L4, which opens the air line to the output filter and head-
wheel air valve on the rear of the tape -transport panel. The output filter
removes water from the air system prior to supplying the air to the head-
wheel motor. Accumulated water is retained in the filter and the hose that
runs to drain solenoid valve 1L3 on the lower shelf of the machine. When
the machine operation stops, the solenoid is de- energized and permits the
trapped water to flow into a pan below the solenoid ( same pan is used for
solenoid 1L2) .
The compressor motor is energized by air switch 1S2 and air -bearing
relay 1K6 (not shown) , which applies operating power to the motor.
Vacuum -Guide Air- Cylinder Supply -Air from the tank also is fed to
solenoid valve 1L5, which is located on a rear-frame shelf behind the
machine. An adjustment valve attached to the solenoid permits the air
pressure to be adjusted to the correct level to set the operating speed of
the vacuum -guide air cylinder. This is accomplished with a needle-valve
control that is adjusted to barely allow air to pass. This needle valve should
never be fully closed, since air passage to the headwheel would then be
blocked. The output of solenoid valve 1L5 is supplied to the vacuum -
guide air cylinder, which positions the vacuum guide and causes tape
penetration by the headwheel according to the level determined by the lead
screw on the vacuum -guide positioner assembly. The vacuum -guide air
cylinder is located behind the tape -transport panel, near the rear of the
headwheel panel.
Cooling Blowers and Air Filter
The motor that drives the low- pressure air and vacuum pump also drives
a blower which is mounted directly behind the air pump. This blower pro-
vides forced -air ventilation to the modules below the tape- transport panel
and to the voltage regulator on the lower shelf of the machine. Air enters
the machine from the bottom shelf, through two flat rectangular filters
mounted behind the removable front cover of the lower shelf of the
machine. The filters are made with multiple layers of slitted and expanded
aluminum which is impregnated with adhesive to trap dust and dirt.
Another cooling blower is provided for the record -amplifier module and
associated filter. This blower is located adjacent to the headwheel blower,
behind the tape -transport panel and directly above the record electrical
filter.
Both blowers operate at all times that operating power is applied to the
machine. If the air supply fails for any reason, an associated switch auto-
matically places the machine in the stop mode. The passage of cooling air
through the blower air vent positions the fin attached to the switch arm
and holds the switch closed. If cooling air stops, the switch returns to its
normally open position and opens the operating power circuit.
The headwheel blower is used to cool the headwheel motor and to keep it
clean. The blower motor operates with ball bearings and is permanently
lubricated. The blower is a multistage type having an oil-wetted wire filter
in a single assembly.
Openings around the headwheel motor allow cooling air to be drawn in
and circulate around the motor. The air then is drawn through a flexible
pipe attached to the headwheel panel, and into the headwheel- blower
filter. Dust and lint are collected in the rotating filter before the air is dis-
charged around the blower motor.
4 -3. TRANSPORT CONTROL CIRCUITRY

Early tape systems contained banks of relays for the rather complex
control functions. In more recent systems, many of these relay control
circuits have been replaced with solid -state flip-flops and a diode matrix.
A simplified block diagram illustrating the basic components of the
mode control system is shown in Fig. 4 -5. The mode control system consists
essentially of the following:
1. Momentary-contact, illuminated mode -selector push buttons on the
control panels.
2. Ten two- transistor binary flip -flop circuits, or "mode" modules, (one
for each mode) and five "driver" modules.
3. A diode matrix which couples the output buses of the modules to the
loads (localized control circuits) throughout the recorder.
4. Unlatch circuits which prevent the recorder from being in two modes
at the same time (except audio and cue record modes) .
5. Inhibit circuits which prevent direct transition between various

modes.
As an example of the basic operation of the mode control system, as-
sume that the recorder is in the stop mode and it is desired to enter the
Fig. 4 -5. Basic vtr control system.

master record mode. When the operator presses the MASTER button, a
momentary- contact switch triggers the master mode module, which latches
into the "on" condition and thereby energizes (grounds) a bus called the
master record bus.
The record bus, along with the output buses of the other mode modules,
is connected to the input of a diode matrix. This matrix may be represented
as two sets of buses crossing each other at right angles but not directly con-
nected to each other. In Fig. 4 -5, the input buses from the mode modules
are shown vertically, and the output buses are shown horizontally. Each
output bus goes to a load or group of loads performing related functions,
such as control of the headwheel motor, capstan motor, and capstan roller.
The loads consist generally of transistor switches, relays, or solenoids.
Where a solenoid is used, the output bus goes to a driver, or power -
amplifier, module which actuates the solenoid when the bus is grounded.
The driver modules are identical to the mode modules, but omission of a
feedback resistor in the external circuit causes them to act as amplifiers
instead of flip -flops.
The record bus is connected by an individual diode to each of the output
buses which perform a function required for entry into the master record
mode. Each diode is oriented to conduct when the record bus is grounded
and to be cut off when the bus is ungrounded. Similarly, the other input
buses are connected to their related output buses by diodes. Thus, when a
particular input bus is grounded, all related output buses become grounded
(through the diodes) , and all loads required for entry into the desired mode
are energized. This arrangement constitutes a simple, reliable control system
which uses no moving parts except the momentary- contact push buttons
which initiate the action.
When the machine is in the master record mode, it is undesirable to
permit direct entry into certain other modes (such as setup or standby)
without first going into the stop mode. To protect the machine from an
error by the operator, unwanted transitions are inhibited automatically by
transistor switches actuated either by the input buses to the diode matrix
or by one of the related output buses. These inhibit circuits apply a voltage
to the mode modules which control the unwanted modes. This voltage pre-
vents the machine from entering the undesired modes even if the operator
pushes the button requesting one of these modes.
When the recorder is in one mode ( such as master record) and the
operator presses the button of another mode which is not inhibited ( such
as the play mode) , the first mode module must be turned off to prevent the
machine from being in two modes simultaneously. To accomplish this, the
output of each mode module (except the audio and cue modules) is con-
nected to a bus, called the unlatch control bus, which goes to a transistor
switch or unlatch amplifier. The output of the unlatch amplifier goes to
another bus, called the unlatch bus, which is connected to the inputs of all
mode modules. When the unit is in master record, a voltage is applied
from the output of the master mode module to the unlatch control bus.
However, this voltage is insufficient to actuate the unlatch amplifier. If the
PLAY button is then pressed, the play mode module increases the voltage
on the unlatch control bus to a level sufficient to actuate the unlatch
amplifier. The unlatch amplifier then applies a voltage to the unlatch bus,
and this voltage turns off the master module. The play module remains on
because the trigger voltage overrides the unlatch voltage. Since only one
mode module is on, the voltage on the unlatch control bus decreases to the
cutoff level, and the unlatch amplifier goes off until the next mode is
selected.
The mode control systems for both Ampex and RCA also contain a tape -
motion detector and play control system that permits safe direct transition
from the wind mode to the play mode without having to actuate the STOP
button. Overstressing of the tape is automatically prevented by delaying
contact of the capstan roller with the tape until after the tape has come to
rest. This circuit also prevents transition from the wind mode to the setup
or record mode while the tape is in motion. These features are particularly
desirable in remote or automatic operation where it is not convenient to
ascertain when the tape has come to rest after winding.
EXERCISES
Q4 -1. Give the sequence of fixed heads (A) across the top of the tape and (B)
across the bottom of the tape.
Q4 -2. What determines head -to -tape pressure?
Q4 -3. What controls tape speed during record or playback operations?
Q4 -4. What is the function of the tension arms?
Q4 -5. What is the function of the tape lifter?
Q4 -6. What elements are supplied by the low- pressure air system?
Q4 -7. What elements are supplied by the high -pressure air system?
Q4 -8. In Fig. 4 -5, if the SETUP and MASTER RECORD push buttons were
depressed simultaneously, in what mode would the system be placed?
CHAPTER 5
The Video and FM System
This chapter will treat the entire video -fm system, bypassing circuitry
more appropriately termed "signal processing." Fig. 5 -1 shows the functions
to be studied in this chapter, while Fig. 6 -1 ( Chapter 6) illustrates the
signal -processing subsystem.
5 -1. FUNCTIONAL DESCRIPTION

There are only minor variations in the terminology for the functions
shown in Fig. 5 -1 for Ampex and RCA. The basic functions of each block
follow.
Video Input: This unit provides for proper termination of the video
line, and it includes a switch allowing unity or variable gain and chroma
controls. The video input unit employs filtering to remove spurious high
frequencies, and it provides pre- emphasis of the video signal to improve
the signal -to -noise ratio. The pre- emphasized video is fed to the modulator.
An amplified video signal from a stage prior to filtering and pre- emphasis
is fed to the modulator afc for the purpose of establishing a black reference
from the signal itself.
Modulator: This section clamps pre- emphasized video at black level to
feed a capacitance- diode -controlled heterodyne -type modulator. The output
is therefore fm.
Modulator AFC: This unit provides precise control of the black -level
frequency of the fm modulator in accord with a crystal -controlled refer-
ence frequency from the fm reference module.
FM Reference (Recording Standards) This module provides the crystal -
:
controlled reference frequency for the modulator afc. It also provides a

white reference frequency keyed into the vertical -blanking interval of the
signal for a check of fm deviation. This white pulse is removed prior to the
video output of the system.
89
Record Equalizer: The record equalizer provides compensation of the

record drive signal so that constant current in the video head is maintained
over the fm passband.
Record Amplifiers: The output level from the record equalizer is in-
creased to a value sufficient for recording on tape. There is one amplifier
for each of the four heads.
Playback Functions
Sync -Channel Output Post-

Video Input
Video -Channel Output Emphasis
Record
Functions
i
HVideo Input Video Filter
l i
Video
Modulator
Demodulator
Clamps IFM
i
Modulator
AFC
r- FM Filter
t r--r ' i
w FM
Reference
I _ FM Equalizer
RCA
Only
Record Sy c
FM Switcher
Equalizer Demodulator
Record Headwheel
tilt
Playback
Amplifiers Panel Amplifiers
P" Red/ Playback

Head Transfer Preamp
Relays
Fig. 5-1. Simplified block diagram of typical video and fm system.
Playback Amplifiers: These amplifiers provide agc, head -resonance

compensation, and aperture compensation for correcting the playback
characteristic. There is one amplifier for each of the head channels.
THE VIDEO AND FM SYSTEM 91
FM Switcher: This unit switches between heads during playback, con-

necting the head scanning the tape to the output. It includes separate
switching for two output channels, one for the video ( picture) and one for
the sync.
Sync Demodulator: The fm signal is limited and demodulated for the
sync channel.
FM Equalizer: This circuit provides additional aperture compensation
for the overall playback equalization characteristic.
FM Filter: This filter provides the precisely controlled overall response
characteristic required for optimum signal -to -noise ratio and frequency
response of the tape playback system.
Video Demodulator: The fm signal is amplitude -limited and demodu-
lated to push -pull video.
Video Filter: This section provides the low -pass filter characteristic
after demodulation required for optimum noise, moire, and frequency
response. Appropriate filters are selected for each fm standard (low -band
monochrome, low -band color, or high -band) .
Post -Emphasis: This unit includes the necessary post- emphasis character-
istic for the demodulated video, and it also provides switching- transient
suppression and video line- output functions.
The video and fm system performs two major functions. These are:
(A) to record incoming video information on storage tape by means of
quadrature heads and (B) to retrieve video information from tape storage
(through the same heads) for playback.
A separate group of modules is used for each of the two functions.
Modules in the record function convert the incoming video to fm and
supply this fm signal through contacts on the record position of relays in
the fm preamplifier module to the quadrature heads on the headwheel.
During playback, fm is extracted from the tape by the heads, amplified
through the fm preamplifier, and supplied to the group of modules per-
forming the playback function for conversion back to video for further
use in the machine.
Special circuitry is included in the video and fm system to permit check-
out and adjustments in the system without the use of the quadrature
heads. This is the electronic -to- electronic (E -E) circuit, which connects
the record and playback functions back-to -back for setup adjustments.
NOTE: Fig. 5 -1 should be referred to often during the remainder of this
chapter to keep in mind the sequence of operations.
Sections 5 -2 and 5 -3 will consider the basic modulation-demodulation
process as typified by earlier (low -band) systems such as the Ampex
1000C and the RCA TRT -1A. A number of these systems are still in daily
use. Then, Section 5 -4 will consider the important refinements of the
combined modulation-demodulation process in most recent systems, typified
by the Ampex VR -2000 and AVR -1, and the RCA TR -70.
5 -2. BASICS OF THE MODULATION PROCESS
The basic function of the modulation process is illustrated in Fig. 5 -2.

The video input (Fig. 5 -2A) is pre- emphasized (Fig. 5 -2B) , to improve
the signal -to -noise ratio, and is used to frequency -modulate a carrier (Fig.
5 -2C) before being applied to the rotating video heads.
The frequency-modulated carrier permits a carrier frequency only
slightly higher than the highest video frequency to be used without objec-
tionable nonlinearity caused by the magnetic medium. The reproduction
of the low-frequency portion of the video band is excellent, even though
(A) Input. (B) Pre -emphasized.
(C) Fm ri output.
Fig. 5 -2. Basic modulator action.
the fm carrier is in a region of poor magnetic performance. Reproduction

of the higher video frequencies is excellent because the lower sidebands
that they produce fall into a frequency band of high magnetic performance
(Fig. 5 -3) .
The block diagram of the RCA modulator (Fig. 5 -4) will serve to
illustrate a typical modulation process. The incoming video signal is ap-
plied to the deviation -control potentiometer. From the center arm of the
potentiometer, the video signal is applied to a pre- emphasis network to
improve the signal -to -noise ratio. The output of the pre-emphasis network
is applied simultaneously to a video amplifier and a sync- stretch amplifier.
The sync-stretch stage provides nonlinear amplification which stretches
the sync region and compresses the white region of the signal.
The video amplifier is a conventional stage with no bypass capacitor
connected across the cathode resistors. Thus, it provides considerable
degeneration which reduces the stage gain but stabilizes the operation of
the amplifier. The video output signal is taken from a low- impedance
cathode circuit and capacitively coupled to the control grid of the re-
Carrier'
Freq
- -
Sidebands of 5-MHz Carrier
,-t--
I
-L
Deviation Range
1.4 2 .5 45 4.3 1 5 6I8

IMHz) I
1
I I
I
I I RF Carrier
I I I
Video
13.6 j 3
I
2.5
I
1.5
I
0.5
I
l
r
I I
MHzI 4 j
I (
II I
I I
I
I I
Sync Tip Blanking White Peak
Fig. 5 -3. Carrier deviation and sideband energy distribution for a

4 -MHz video input signal.
actance tube. The purpose of the reactance -tube modulator is to frequency-

modulate a radio-frequency oscillator. The modulator accomplishes this by
changing the amplitude variations of the input signal into a varying re-
actance which is connected across the tank circuit of the oscillator. The
plate voltage of the reactance tube is held constant by a voltage regulator,
to insure that the clamp bias voltage applied to the grid of this tube holds
the carrier frequency limit constant for the tip of sync.
The fm output signal of the 46 -MHz oscillator is applied to the control -
grid circuit of a mixer stage. Here, it is heterodyned with the master-
Video Mono
Input SIA 4.0 dB Video Reactance FM Osc
Amplifier Tube Nom 46MHz
y
Deviation
Color;
\
9.54 dB
Pre- Emphasis
Control s Networks
Sync Fixed Osc

White Pulse White Sync Mixer
Stretch Clamp
51 MHz
Amplifier Tip Tip
Amplifier
S2A
Clamp Pulses
on Sync Tips
Meter
FM
Pulse Gate Amplifier
H J
White
SIB Tip -Sync Tip SIC
Mono
FM Output
Reference Tank Circuits
IApprox
for Frequency and
1 -8 MHz)
Deviation Indicator
Fig. 5 -4. Simplified block diagram of RCA TRT -1 A modulator.

oscillator signal, which is injected into the suppressor -grid circuit (Fig.
5 -5) The frequency of the master oscillator is 51 MHz, but it is not critical.
.
The operation of the master oscillator is the same as that of the fm oscil-
lator. The coupling between fm- oscillator stages (Fig. 5 -5) is through
common cathode impedance L8 and plate -to -grid coupling capacitor C17.
There are two heterodyne signals from the mixer; one is the sum of the
master -oscillator and fm- oscillator frequencies, and the other is the differ-
ence between the two oscillator frequencies. The output of the mixer is
applied to a two -stage amplifier (Fig. 5 -4) . Inductors in the control -grid
Video Input
Reactance Tube
22 pF
C17 p
1000
From Driver
Stage) n ?y
Nominal
Coupling Capacitor Common Center
Brought to Reference - Impedance - Frequency
Voltage During = 46MHz
Sync -Tip Time
Conducts - FM to
Cutoff - Driven Clamper

Voltage
Regulator
Output
Stages
From Sync
Stretch -pH 51 MHz
From
Tube
Fixed
Cut Off by Grid Osc
Current Between
Sync Pulses
?RO
Fig. 5 -5. Simplified diagram of modulator, clamp, and mixer.
circuits of these amplifiers provide high- frequency peaking to keep their

frequency response flat. Also, these coils prevent unwanted oscillator -
frequency signals and their sum from getting into the fm output. The
amplifier output signal is frequency -modulated, with a center frequency of
approximately 6 MHz ( midway between 4 and 8 MHz for a signal with
50 percent duty cycle). A portion of the output voltage is applied to the
cathode of the pulse-gate tube (Fig. 5 -4) . This tube is biased at plate -
current cutoff, except when its control grid receives positive-going pulses.
A portion of the input video signal is applied to the white -pulse am-
plifier. Both sections of this stage are biased so as to stretch the white
information and compress the sync. When switch S2 is placed in the
White -Tip position, one of the two rf reference circuits ( depending on the
position of switch S1) is energized during the whitest part of the picture.
The rf voltage generated by the reference circuit is rectified by a germanuim
diode, and the amount of current is indicated on the meter. Maximum de-
flection of the meter is obtained when the frequency of the carrier output
equals that of the reference circuit selected. When S2 is in the Sync -Tip
position, positive sync -tip pulses force the gate tube into conduction,
thereby permitting the fm carrier frequency at the sync tips to be checked
with the other two reference circuits.
The Ampex 1000 -C modulator employs slightly different oscillator fre-
quencies, but the principle of operation (heterodyne to obtain carrier
frequency) is the same for systems following the original Ampex Model
1000 -A. The Ampex 1000 -C includes the Ampex standard monochrome
(somewhat higher pre- emphasis than RCA) and color networks, as well as
a considerably higher pre- emphasis termed Ampex Master Video Equaliza-
tion (AMVE) -any one is available by means of a push- button switch.
In addition, the equalizer components are plug -in units.
The Ampex 1000 -C modulator incorporates built -in crystal- controlled
test signal generators. Since carrier and deviation are set differently for
color than for monochrome, the modulator provides separate presetting
adjustments for the two. The recorder can be switched, either locally or
remotely, for either color or monochrome operation. The switchover
automatically actuates the correct preset carrier and deviation values in the
modulator.
The following crystal -controlled test frequencies may be selected by
push buttons on the Ampex modulator:
5.0 MHz-Monochrome blanking level
5.5 MHz-Color sync tip
6.5 MHz-Color peak white
6.8 MHz- Monochrome peak white
Two additional positions are also available on the Ampex. They allow the
user to plug in units for any special crystal-controlled frequencies.
The Ampex 1000 -C modulator employs a carrier frequency clamped at
the blanking level during the back-porch interval (Fig. 5 -6) The sync is
.
separated from the composite signal (waveform A) and differentiated

(waveform B) to form leading -edge and trailing-edge spikes. The trailing
edge is used to form a clamp pulse (waveform C) , which drives the
clamper at the blanking level on the back porch.
5 -3. DEMODULATION BASICS

The purpose of the demodulator is to receive the fm carrier from either
the switcher (playback mode) or the modulator (record mode) and extract
the video information from the carrier signal. A block diagram of the RCA
TRT -1A demodulator unit is shown in Fig. 5 -7. It illustrates the necessary
processes for wideband demodulation. During the playback mode of opera-
tion, relay K1 is unenergized, allowing the input signal from the 2 X 1
switcher to appear at the grid of amplifier stage VIA. Simultaneously, it

grounds the input from the modulator chassis. In the record mode of opera-
tion, 24 volts dc is applied to K1. This energizes the relay and reverses the
foregoing situation. The input signal from the modulator is now applied
to the grid of amplifier VIA while the input from the 2 X 1 switcher is
grounded.
Horiz Blanking
Z^
Sync
,¡Front Back
-H
Porch Porch
Fig. 5 -6. Derivation of clamp pulse

from sync for back -porch clamping.
Stage VIA acts as a straightforward fm amplifier. A coil and capacitor

are employed in the plate circuit to obtain the proper frequency response.
The amplified fm signal is then coupled to the control grid of phase splitter
V1B. (Stages VIA and V 1B are shown as a single block in Fig. 5-7.)
Two fm signals 180° out of phase are taken from the plate and cathode
circuits of V1B and fed to the control grids of push -pull amplifiers V2 and
V3. Approximately 50 dB of limiting is achieved by diode clippers to
minimize dropouts and other effects of amplitude modulation.
The balanced output voltage is fed into four successive stages of push-
pull amplification, each consisting of matched pairs of tubes. Push -pull
amplifiers V12 and V13 raise the level of the clipped fm signal to 23 volts
(peak -to- peak). The signal from the plate circuit of V13 is fed directly to
a control grid of phase splitter V14, while the signal appearing in the plate
Fig. 5 -7. Block diagram of the RCA TRT -1A demodulator.

Fig. 5 -8. Simplified diagram of duty-cycle (pulse-width gating) section of

RCA TRT -1A demodulator.
circuit of V12 passes through delay line DL1 before being fed to the
opposite control grid of V14. Delay line DL1 shifts the signal phase 180°
at 4.0 MHz, thus enabling the signals which appear at the control grids of
V14 to be in phase at that frequency.
Each section of phase splitter V14 supplies two signals 180° out of phase
to a pair of matched 6BN6 tubes (V15 and V16) comprising the dis-
criminator stage (Fig. 5 -8) The signal appearing in the plate circuit of
.
V14A is applied to the suppressor grid of discriminator tube V15, while

the signal appearing in the cathode circuit of V14A is fed to the suppressor
grid of discriminator tube V16. The signal appearing at the plate of V14B
is fed to the control grid of V15, while the signal appearing at the cathode
of V14B is fed to the control grid of V16.
Input Frequency Slightly Above 4 MHz Input Frequency Slightly Below 8 MHz
Und GriSignal I
on Gri d I
H-0.125 vs-
Delayed and Inverted
Signal on Grid 1 JII
lil
II
0.125 vs Plate -Voltage Waveform 0.125 ps-

il
Screen -Voltage
Waveform
U U (Inverse to Plate Waveform I
Fig. 5 -9. Demodulator waveforms for one of two gate tubes.
Tubes V15 and V16 may be considered as pulse -width gating tubes
because of their ability to produce pulses of varying width and constant
amplitude. The characteristics of these tubes are such that plate current will
flow only during the time a positive voltage appears at the control and
suppressor grids simultaneously. Therefore, the width of the negative pulses
appearing in the plate circuits of V15 and V16 is equal to the overlap
width of the positive pulses applied to their signal grids (Fig. 5 -9) . The
widest negative plate pulse occurs when the signals are in phase, and the
narrowest plate pulse occurs when the signals are oppositely phased. If the
suppressor grid is negative, the current is to the screen; if it is positive, the
current is to the plate. The duration of plate current in either tube is
directly proportional to the phase difference between the signals applied to
the control and suppressor grids. The net result is that the 6BN6 dis-
criminator amplitude-versus-frequency slopes are inherently linear.
Pulses appearing in the plate circuit of V15 and V16 do not occur
simultaneously, but are uniformly interspaced in midpositions because of
the phase -splitter action of V14. Thus, if the sets of pulses are well bal-
l0o TELEVISION BROADCASTING
anced, the combined gate output consists only of video information carried
by even harmonics of the fm carrier.
The characteristics of the 6BN6 tube are such that the sum of the plate
and accelerator currents tends to remain constant and independent of both
signal -grid potentials. Thus, a signal current in the plate circuit must be
accompanied by a nearly equal signal current of opposite polarity in the
accelerator circuit, thereby effectively providing a push -pull output from
each gate. The push -pull output signals are fed to cathode -coupled adder
stage V17B (Fig. 5 -10) . Note that the varying duty cycle from 4 MHz
to 8 MHz results in a dc level which represents the instantaneous fre-
quency -effectively converting frequency change to dc change.
Video -adder stage V 17B receives video information from the plate and
accelerator circuits of discriminators V15 and V16. This information is
added by V17B and appears at the plate. The video output is then fed into
a 10- section low -pass filter (DL2) which removes the carrier harmonics
and unbalanced fundamental carrier frequency above 5.2 MHz. When the
deviation of the fm carrier does not occur lower than 5.2 MHz, as in color
operation, this filter adequately removes small unbalanced carrier com-
ponents that would degrade the video signal. However, when deviation
occurs down to 4.3 MHz, with blanking at 5.0 MHz as in monochrome
operation, discriminator balance requires more careful adjustment to keep
blanking free of residual carrier not removed by DL2.
The video signal from filter DL2 is fed into video amplifier V18 (Fig.
5 -7) . In the plate circuit of this tube is a de- emphasis network which
matches the pre- emphasis circuit located on the modulator chassis. These
circuits serve to improve the signal -to -noise ratio.
Slightly Above 4MHz Slightly Below 8MHz
F-0.125 us- Upper -Plate 0.125 us

Voltage Waveform
1
®
i
0.125 lis Lower -Plate

Voltage Waveform
i
, l
O -
Sum of Both
Duty Cycle Becomes Z ro at 4 MHz

Plate Waveforms J L-1.
Duty Cycle Becomes 100% at 8MHz
O
U
Upper -Screen
Voltage Waveform J L
Lower -Screen
U Voltage Waveform
Fig. 5
®
Sum of Both
Screen Waveforms_ - .-
-10. Demodulator waveforms for both gate tubes together.
Switch S2 also opens or shorts a lead fed to the chroma-processor chassis

(color equipment). This lead is in series with a circuit which operates a
relay on the processor chassis. Its purpose is to make certain that the relay
is not triggered by false color bursts when the tape recorder is operated in
the monochrome mode.
A series -resonant circuit is located in the cathode circuit of V18. This
circuit may be switched out, to compensate for small high- frequency losses
inherent in the modulation and demodulation processes, by placing normal-
rolloff switch S1 in the rolloff position. This position may be desirable
to reduce excessive noise for either mode of operation of the tape recorder.
However, when SI is placed in the rolloff position, signal resolution is lost
with the noise.
Output stage V19 operates as a conventional video amplifier. The final
video output signal is taken from its cathode and fed to the 2 x 1 switcher
chassis through output jack J4.
The discriminator slopes may be seen readily by applying a 1 -volt peak -
to -peak sweep signal to input jack J2 and observing the waveform at output
jack J4. Fig. 5 -11 indicates the ideal waveform (solid line) obtained with
the demodulator well balanced and switch S2 (color- monochrome) in the
monochrome position. The single -line slope from 4 to 8 MHz is the dis-
criminator slope used for demodulation. The large envelope at the left is the
second harmonic which appears to be cut off at 2.6 MHz but is actually cut
off at 5.2 MHz, the cutoff point of DL2.
A simplified block diagram of the Ampex 1000 -A demodulator is shown
in Fig. 5 -12. During replay, the signal enters the demodulator from the
switcher assembly and is fed to a 60 -dB limiter strip. The limiter strip, in
turn, drives a delay-line type of demodulator. The delay is chosen to repre-
4.3 6.8
Monochrome
Deviation Range
E
Q
\ i
Cola Deviation Range
2 2.6
\ \\ \
\\
4
/ / //
5 2
5 5
6
6.5
MHz
Filter
Cu off
Fig. 5 -11. Video -sweep response of demodulator.
sent a 90° phase shift at the carrier frequency of 5 MHz (or 0.05 micro-
second) . The signal is split into two paths at the output of the last limiter
stage. One path proceeds directly to the grid of the adder tube, while the
other goes through the delay line before being fed to the adder stage. The
diagrams in Fig. 5 -13 illustrate what happens when the carrier frequency,
a lower frequency, and a higher frequency are passed through this circuit.
The resultant amplitude varies as a function of frequency; therefore, fm
demodulation is accomplished.
The am signal thus produced (which still contains an fm component) is
detected by a pair of diodes driven by a transformer, as shown in Fig. 5 -12.
The transformer acts as a phase splitter to provide the signal for full -wave
rectification, which is employed to achieve full -time signal recovery and
to cancel the unwanted second harmonics of the rf signal. After further
amplification, a low -pass filter eliminates all of the double frequencies
developed in the full -wave rectifier diodes and most of the carrier frequency
(filter cutoff is approximately 4.5 MHz) .
From 60 -d B Cathode 0.05 us Phase

Adder
Switcher Limiters Follower Delay Line Inverter
Filters Out
Fundamental of
Carrier
Detector
Video Low -Pass Balance
Amp
Out Filter Balances Out 2nd
Harmonic of Carrier
Fig. 5 -12. Simplified block diagram of Ampex demodulator.
5 -4. REFINEMENTS IN MODERN

MODULATION -DEMODULATION SYSTEMS
To get a good visualization of one of the basic problems of video tape
recording, it is necessary to consider the modulation and demodulation
t
1
r1
Carrier Frequency Higher Than Carrier Lower Than Carrier
Fig. 5 -13. Vector analysis of delay -line demodulator.

processes together. Study Fig. 5 -14 carefully. Fig. 5 -14A is a plot of signal
versus frequency of the video signal itself and the carrier and deviation
frequencies of low -band and high -band recorders.
The carrier frequency is deviated by the varying amplitudes of discrete
frequencies in the video signal. This produces upper and lower sidebands
with energies particularly strong at the second-order and third -order side-
bands (first and second upper and lower sidebands) .
Video Law Band High Band
3.58 4.2
.Color_
5.5 6 5
Mono Mono & Color

Frequency (MHz) 4 3 6.8 7.06 10
(A) Vtr carrier and deviation frequencies.
O O O O
I I I
I
I
I
-
I I I I
I, 1. 1 I
2 -1 0 1.8 2.4 3 6 6 0 90
Color Carri r Freq Carri r Freq
Foldover I
t Subcarrier (Low Band) (High Band)
Frequency (MHz) II
O 1st Lower Sideband, Law -Band Color

O 2nd Lower Sideband, Law -Band Color
O Folded Sideband, Spurious Signal, Low -Band Color
(D 1st Lower Sideband, High -Band
O 2nd Lower Sideband, High Band
(B) Carrier frequencies vs color sidebands.

Fig. 5 -14. Frequencies in video recording.
In low -band operation (for monochrome) even though the lower devia-
tion edge is adjacent to the highest- frequency video input component,
little effect is noticed in normal picture content. You will see the spurious
effect of this on the high- frequency wedges of a monoscope pattern, or on
the higher multiburst frequencies. It appears as a "beat" or moire inter-
ference on this type of test signal. But the monochrome program signal
has very low energy at the upper end of the passband, so very little sub-
jective interference can be noticed with this type of signal.
Now consider the low -band color operation. Here there is a discrete fre-
quency of high energy (the color subcarrier) beating with the instanta-
neous fm carrier. As deviation is increased, the sidebands increase rapidly
in amplitude. Hence it is necessary to reduce the deviation in low -band

color operation as compared with low -band monochrome operation.
Now refer to Fig. 5 -14B. Assume the average carrier frequency (50-
percent APL) to be 6 MHz for low-band color. Also assume that the color
subcarrier frequency is 3.6 MHz instead of the actual 3.579545 MHz. This
allows for easy computation to make a point.
The first lower sideband ( for the color frequency) occurs at 6 3.6 -
-
= 2.4 MHz. The second lower sideband then is 2.4 3.6 = -1.2 MHz.
Thus the excess energy which attempts to occur below zero frequency
"folds over" as a spurious signal at plus 1.2 MHz. The energy distribution
becomes rather broad around this frequency, and a strong moire pattern
(interference pattern) appears in the color information. This, of course,
increases in intensity with degree of saturation (amplitude) of the color
signal. Hence it is very pronounced on color bars.
Since the low -band video heads had an upper "shelf" of around 7.2 MHz,
the upper sidebands were discarded and corresponding standards were
adopted. Modern low -band recorders used strictly for monochrome still
reflect this condition. But modern high -band recorders, which can also be
switched to low -band operation, are different. The first and second upper
sidebands are now accommodated, and a distinct improvement in video
signal processing is possible. Thus, you will see an improvement even in
low -band color quality on modern systems.
At this point, improved video -tape modulation methods will be discussed.
The terms "heterodyne" and "dual heterodyne" will be used to describe the
function. To avoid any possible confusion, it is necessary to clarify the
difference between modern heterodyne modulators and the older "hetero-
dyne color" systems used by both Ampex and RCA.
In these original attempts at color video tape, the station subcarrier-
generator signal was used as a reference against which to measure "jitter"
of the subcarrier component coming from the tape playback. This com-
bined signal was changed in frequency (heterodyned) to obtain a signal
corresponding to the jitter. The jitter error was cancelled, and then the
frequency was heterodyned back to normal. In this system, no definite
relationship exists in the final output signal between the trailing edge of
sync and the frequency of the subcarrier. Hence there is no "dot interlace,"
and the signal is "nonphased" to local sync.
The "heterodyne color" technique is now only a matter of historical
interest and is not treated here. However, it is the reason for the special
mode of operation provided on modern recorders by means of nonphased
color (NPC) circuitry.
The reactance -tube heterodyne modulator described in Section 5 -2
(which did provide standard dot -interlaced color) has now been replaced
by a "dual heterodyne" modulation technique. This method has a much less
severe limitation on the differential -gain, axis-shift, and phase -shift char-
acteristics than that of the reactance control circuit.
All circuits are now almost strictly solid state. Exceptions are in the first
stage of the video -head playback preamplifier (where nuvistors are used)
and in the video -head record drivers.
In solid -state modulators, the voltage -variable diode capacitor is used.
A review of this application in video modulators is presented in Section 5 -5.
Since this device has a nonlinear capacitance voltage relationship, the
result is about the same as that obtained with the average reactance circuit
(Fig. 5 -15).
F2 -F1= Fout
F2
Variable
Fin
F1
Fixed
Fout
Fig. 5 -15. Typical characteristics of reactance circuit.
The new dual heterodyne technique (Fig. 5 -16) overcomes this limita-
tion. One oscillator operates at about 100 MHz and the other operates at
about 108 MHz. Under modulation, they are deviated in opposite direc-
tions, with the result that cancellation of amplitude- frequency nonlinearity
takes place. The actual modulator frequencies depend on the particular
manufacturer and model number, but the principle is the same.
The difference frequency (modulated carrier) is obtained from the low -
pass filter (Fig. 5 -16). Video feed -through and even-order harmonics are
180° out of phase at the output, resulting in improved suppression. The
overall modulation transfer characteristic also results in about a 2-to-1
F3
Mod Carrier
Fig. 5 -16. Dual heterodyne modulation.

increase in dynamic range. The main point to keep in mind is that the
variable- reactance diodes are arranged in complementary fashion so that
oscillator frequencies F1 and F2 change in opposite directions with respect
to the input signal.
5 -5. MODULATION CIRCUITRY

We will first review the characteristics of the voltage -variable- capacitance
diode. The width of the barrier in a pn junction is influenced by the voltage
across the junction. Since an electrostatic field is thus produced, there is
the equivalent of two plates of a capacitor in which the spacing between
the plates is governed by the voltage ( and current) of the junction.
Figs. 5 -17A and 5 -17B illustrate this effect. Anything that increases the
width of the pn junction barrier is equivalent to spreading apart the plates
of the capacitor, resulting in less capacitance. As the reverse bias is de-
creased, the junction capacitance increases.
Electric Field Barrier Smaller
unction C
Junction
Electric
Field Barrier
"F-
P r
Junction III
Capacitance
(A) Zero bias. (B) Reverse bias.

Fig. 5 -17. Voltage -variable- capacitance diode.
Fig. 5 -18A illustrates the junction- capacitance curve typical of nearly all
signal diodes. Note that, in the reverse -voltage region of operation, the
capacitance change is quite small over a 100 -volt excursion. By comparison,
Fig. 5 -18B is an example (only) of the curve for one type of variable -
capacitance diode. There are many different types of variable-capacitance
diodes with different sensitivites in capacitance per volt of change. All of
these diodes typically have a nonlinear relationship between capacitance
and voltage change, as shown by this graph.
Some common circuits employ a voltage-variable diode capacitor, which
may be known as a 'variable- capacitance diode," "varicap," or "varactor."
This is a diode which is manufactured to maximize certain physical and
chemical features affecting junction capacitance. Fig. 5 -19 shows three
symbols for this type of diode that may be found on schematic diagrams.
2.0 100
1.9 80
a 60
U
2, 40
16 20
0 01 0 1 1 0 10 100 2 3 5 7 10
Reverse Voltage Reverse Voltage
(A) Typical small- signal diode. (B) Variable - capacitance diode.

Fig. 5 -18. Variation of junction capacitance with voltage.
Like the zener diode, the variable -capacitance diode is always operated
in the reverse -biased condition. Capacitance decreases with increased re-
verse bias; that is, the capacitance varies inversely with the reverse bias
voltage. The capacitance- versus-voltage relationship is nonlinear. The diode
normally has a greater change in capacitance for a given change in applied
voltage than the effective capacitance change in a conventional reactance-
tube circuit.
When the total capacitance used to resonate a tuned circuit is split and
used to form a feedback voltage divider, a Colpitts oscillator results. Fig.
5 -20A illustrates a transistorized Colpitts oscillator, in which Cl and C2
Cathode
Fig.5 -19. Symbols for voltage -

variable- capacitance diodes.
Anode
together set the effective capacitance with which L resonates. Remember

that part of this capacitance is "hidden" in junction capacitance of the
transistor. The important point at this time is to understand the effect of
a capacitance diode on an LC tuned circuit.
Now look at Fig. 5 -20B. This is the same circuit as the one in Fig. 5 -20A,
but with added components. Variable- capacitance diodes X1 and X2 must
remain reverse biased at all times; this bias is supplied through resistors
R1 and R2. Capacitors C3 and C4 are large and have negligible reactance
to, let us say, a video signal. Now assume that a video signal is applied
with sync in the negative polarity. The effective capacitance of X1 and X2
varies with the instantaneous video voltage. Since the diodes are effectively
connected to the tank circuit, the resonant frequency is varied in step with
the instantaneous video voltage variation. Let us see now in which direction
the frequency changes between sync tip and peak white. Since sync is the
most negative signal, the reverse bias is increased. Since capacitance varies
inversely with the reverse bias, capacitance decreases, increasing the output
frequency of the oscillator. The same reasoning tells us that as the signal
goes toward peak white, the frequency decreases. The result is a frequency -
modulated (but nonlinear) output.
(A) Colpitts oscillator. (B) Fm Colpitts oscillator.

Fig. 5-20. Application for voltage -variable -capacitance diodes.
Note that with the same type of transistor and the same polarity of video
input signal, the polarity of X1 and X2 could be reversed, with the reverse
bias now negative rather than positive. In this case, the oscillator would be
caused to decrease in frequency at sync tip and increase in frequency as the
signal swings toward peak white. If the same video signal is fed to both
inputs of the foregoing oscillator circuits, one oscillator increases in fre-
quency as the other decreases. This is a push -pull type of frequency modula-
tion which tends to cancel the nonlinearity of modulation resulting from
the nonlinear voltage-capacitance relationship of voltage -controlled capaci-
tance diodes. When this type of modulation is not used, a nonlinear
amplifier of characteristics inverse to those of the modulator restores
linearity.
Obviously, this type of application is not limited to the Colpitts oscillator.
Nor is it limited to the production of frequency modulation. For example,
this type of diode is used in voltage -controlled delay lines (in place of the
shunt capacitors) as applied to automatic time correction (ATC) in video
tape recorders (Chapter 6) .
A typical video -modulator circuit arrangement for a modern tape system
is shown in simplified block form in Fig. 5 -21. Pre -emphasized video is
supplied through the deviation control, which allows proper setting of the
tc
m LL
fi 1
o.
E
.-L
i ii 1
ALL
o.
E
E i
c)
v
U
Q
Fig. 5 -21. Typical video modulator arrangement for television tape systems.
fm deviation depending on the recording standard used. The amplitude -

controlled signal then goes to a pair of transistors in a complementary -
symmetry emitter -follower stage (CSEF) . The relative outputs from these
two emitter followers are balanced by means of the video balance control
and fed to the two fm oscillators through low -pass filters and across a
clamping arrangement in both the low- frequency and high- frequency oscil-
lator paths.
The low -pass filters are used to prevent rf from feeding back into the
video circuits. The filters also usually provide part of the reactance required
for the oscillator tuned circuits that follow the varactors.
Video in each channel is clamped (either during the back -porch inter-
val or the tip -of -sync interval) to a reference voltage potential. Since this
potential is part of the reverse -biasing voltage on the varactors, it becomes
an influence on the operating frequency of the oscillators. This reference
voltage is dependent on the following:
1. A dc level from the fin reference (record standards) module. This
reference voltage depends on the particular standard selected.
2. The position of the vernier frequency control on the modulator.
3. A fixed basic bias level usually obtained from a zener diode and
resistor network.
4. The afc error pulse amplitude and polarity.
Each oscillator output is limited to remove amplitude modulation that

would cause spurious frequency components in the modulator output. The
output of the low- frequency limiter is amplified and passed through a filter
having a bandpass of 90 to 110 MHz to remove all undesired components
from the low- frequency channel. The output of the high- frequency limiter
is supplied to a buffer amplifier and then to one side of a mixer circuit the
opposite side of which receives the filtered output from the low- frequency
oscillator.
The low- frequency channel drives the mixer in push -pull by means of a
transformer. The high- frequency channel drives the mixer in an unbalanced
manner. Eventually, the high- frequency signal is cancelled by the common -
mode rejection of the following differential amplifier; however, the com-
ponents from the low- frequency channel are not cancelled. A mixer balance
network is included to permit accurate balancing of the mixer, to avoid
generation of spurious frequencies.
Outputs from the mixer are supplied ( actually through low -pass filters
having linear phase -shift characteristics) to the differential amplifier in
push -pull. These signals are combined to form a single -ended output to
feed the line driver. The line driver supplies two outputs. One output feeds
the modulator afc module. The other output is coupled through a network
that compensates for phase -shift characteristics of the bandpass filter in
the low -frequency oscillator channel. The phase -compensated signal is fed
to the record equalizer module, usually through a low -pass filter to remove
residual high- frequency products that have not been cancelled in the
mixer.
Modulator AFC
All modern tape systems employ a means of automatic frequency control
(afc) for the video fm modulator. See Fig. 5 -22A. Pre -emphasized video
is fed to the modulator afc. The fm reference circuitry supplies crystal-
controlled reference frequencies. The output of the modulator is also fed
back to the afc circuitry.
Pre- Emphasized
To
Video
P Modulator Rec
Amps
Mod
(A) Application in system. Video 1.
AFC
FM Ref
(Rec Stds)
Back Porch Sample FM Black Ref n

I
Pulse From Video (From Mod) From FM Ref

(Rec Stds)
Modulator
lus Sampling
Video L F VCO
Delay
Modulator AFC
1
Video H-F VCO ¡-- Cl mp

Pulse
Limiters
11
+
_FL__ Freq Too High
DC Freq Correct
rra 7U Freq Too Low
I(
JI°°
Clamp Pulse
+
J 1 us
Clamp Pulse
U --.1 1- -1.5us
(B) Method of providing afc.

Fig. 5 -22. Modulator afc function.
In Fig. 5 -22B, note that each of the voltage -controlled oscillators (vco)
is returned to a clamping bridge. For the clamping -pulse duration (which
occurs at a 15.75 -kHz rate), each vco is returned to a reference voltage.
Since the effective capacitance of the varactor diodes depends on voltage,
this sets the frequency of the oscillators.
Observe at this time that there is capacitive coupling to each of the
clamped reference returns from point 1.
The afc essentially has three inputs: (1) a back-porch sample pulse
formed from separated sync from the incoming video, (2) the fin output
from the modulator, and (3) a crystal -controlled black-reference frequency
from the fm reference (or record standard) . The sampling pulse is around
3 -µs wide. The clamp pulses are formed from a trailing -edge delay circuit
so that they are delayed from the leading edge of sampling by about 1 µs
with a duration of about 1.5 µs.
White Reference
Low -Band Low -Band High -Band
Reference Mono MHz Color MHz MHz
Sync Tip 4.28 5.50 7.06
Gray Reference
Blanking 5.00 5.79 7.90
Gray (50% APL) 6.00 6.00 9.00

Blanking (Black)
Reference White (Peak) 6.50
6.80 10.00
-Sync -Tip Reference
(A) Relationship to waveform. (B) Frequency chart.

Fig. 5 -23. Reference carrier frequencies.
Note that the reference is at blanking level (blanking frequency of the

vco) and that the modulator output during this interval is being sampled
and compared with the black- reference oscillator frequency. Any difference
between the modulator output frequency and the frequency of the crystal
reference appears as a pulse fed back to point 1 of the modulator. The am-
plitude and polarity of the pulse depend on the degree of frequency differ-
ence and the direction of difference, respectively. The presence of the pulse
changes the clamped reference voltage return of the vco so that the oscil-
lator frequency increases or decreases as required. Note that this reference
is back -porch level; both RCA and Ampex now clamp on the back porch
for the modulator afc reference.
Exactly how does clamping on the back porch affect the modulator?
During normal operations, the vco's are corrected in frequency line by
line at the frequency of the back-porch level (see Fig. 5 -23) .
The other reference frequencies shown in Fig. 5 -23 are available as test
frequencies in most recent systems. Also in these systems, you will find a
"white reference" pulse in the demodulation circuitry; this pulse permits
checking and setting the deviation for proper recording level.
The SMPTE standards for modulation carrier frequencies list an outside

limit of ± 50 kHz. Modern vtr's with afc hold these frequencies well within
± 12 kHz; in fact, warning lights indicate when this limit is exceeded.
FM Reference (Record Standards)
The following description of the fm reference module in the RCA
TR-70 system will serve to illustrate the functional aspects of the circuitry.
The Ampex terminology is "record standards."
The fm reference module ensures that recordings are made according
to established standards. The module is also part of the built -in test equip-
ment in the machine. Three basic purposes are served, as follows:
1. Standards Generator: The standards generator supplies a dc voltage
to the modulator module; the voltage is related to the machine operat-
ing standard and controls the modulation frequency according to the
selected standard. This function also generates crystal -controlled
standard test frequencies to coincide with gray, black, and white sig-
nals according to the selected operating standard.
2. White Insert: This function inserts (by electronic switching) the
selected white test frequency into the vertical -blanking interval for
the E-E mode of operation and during certain tests in the playback
mode.
3. Sweep Generator: The sweep -generator function provides linear
ramp pulses that are supplied to the modulator module during certain
tests and cause that module to function as a sweep generator.
With three distinct functions performed by this module, three signal
paths are required. Each signal path is described below. See Fig. 5 -24
during the following analysis.
A negative 20 volts is supplied from the standards switcher to one of five
inputs on the fm reference module. The particular input supplied is deter-
mined by selection of one of five operating standards. Once inside the
module, the minus 20 volts is fed in parallel to the oscillator control diode
matrix and to one center arm of relay K3, K4, or K5.
Each of the fixed contacts of relays K3, K4, and K5 is connected to a
precision resistor. The minus 20 volts is thus fed through the particular
resistor associated with the selected standard, and this resistor forms a
voltage divider with another precision resistor that is common to all the
selected resistors. The opposite end of the common resistor is permanently
connected to positive 20 volts; therefore, a precisely controlled dc voltage
is made available from the voltage divider for the modulator- module carrier
frequency.
Relays K3, K4, and K5 remain de- energized during normal operation to
provide a black modulation- frequency control level to the modulator
module. However, during tests, a gray modulator output frequency is re-
quired, and the relays become energized by a ground signal originating in
11 1 TELEVISION BROADCASTING
eir
t
S
d
b
t
S
o
Fig. 5 -24. Block diagram of RCA

THE VIDEO AND FM SYSTEM I I
ce
a'
io
CO o C.
O O
o G ?Ñ
Ln
ó Ñ
TR -70 fm reference module.

the fm test panel. This selects a different resistor for the voltage divider
and provides a gray dc control for the modulator.
As mentioned earlier, the minus 20 volts supplied by one of the five
standards is also fed to the oscillator control diode matrix. This matrix
produces control levels which activate specific transistor switches so that
the precise reference frequencies required for the standard selected are
produced by the crystal -controlled oscillators. The correct switches are thus
preset in the gray, black, and white oscillators, so that merely the applica-
tion of power to one or more of the oscillator circuits is required to obtain
the correct gray, black, or white reference frequency from that oscillator.
The particular frequency and the associated transistor switch that is
selected, as well as the position of the TEST SELECT switch on the fm test
panel, are listed in Table 5 -1.
Table 5 -1. Frequency -Standard Selections
Standard
4 5
Test K3 1 2 3 525 625
Select K4 Function 525 625 525 High- High -
Position K5 Activated LB Mono LB Mono LB Color Band Band
Norm Black 5w Q6 Q7 Q8 Q9 Q10
Black Freq 5.0 MHz 5.54 MHz 5.785 MHz 7.9 MHz 7.8 MHz
or Off
White *Sw Q1 Q1 Q2 Q3 Q4
Freq White Freq 6.8 MHz 6.8 MHz 6.5 MHz 10 MHz 9.3 MHz
* *Input On - - - - - -
Gray Sw Q14 Q14 Q14 Q15 Q16
Noise On
Gray Freq 6 MHz 6 MHz 6 MHz 9 MHz 8.5 MHz
* *Res Off - - - - -
°Activated only in E -E mode or in play mode when white insert function is selected.
O*No reference frequency is generated in this mode.
The principles of operation of all frequency generators are similar; thus,

a description of the gray frequency generator (Fig. 5 -25) will suffice.
The gray frequency generator consists of an oscillator (Q17) and three
quartz crystals (M10, M11, and M12). Each crystal is associated with a
transistor switch (Q14, Q15, and Q16, respectively), which activates one
of the crystals on command in a manner described below.
The emitters of the three switching transistors are connected to a com-
mon bus which is bypassed to ground by capacitor C34. This bus is held at
a nearly constant dc level of minus 0.45 volt by resistors R79 and R80.
The bases of the three switching transistors receive their input control sig-
nals from the diode matrix, and are biased by 11K resistors R73, R75, and
R77, which are connected to the plus 20-volt supply.
Fig. 5 -25. Simplified schematic, gray- frequency generator.

In the absence of a negative control signal from the diode matrix, the
transistors are cut off by the positive base bias. This opens the series circuit
from the oscillator through the crystal, the switching transistor, and capaci-
tor C34 to ac ground.
To activate a crystal, the base of the associated transistor switch is driven
negative by an output from the diode matrix; the transistor switch saturates
and connects one side of the crystal to ground through capacitor C34. The
opposite side of the crystal is connected to a common bus by which all
crystals in the generator are coupled ( through capacitor C35) to the base
of oscillator transistor Q17.
A Colpitts configuration is used for the oscillator circuit, with the col-
lector of Q17 connected to ac ground. The selected crystal, which repre-
sents the resonant circuit in the oscillator, is effectively connected between
the base and collector of Q17 because of the saturated condition of its
grounded switching transistor. Oscillation is sustained by feedback caused
by the connection of the emitter to the capactive voltage divider consisting
of C37 and C38.
Diode clipper X20 -X21 limits the base voltage swing so that the output
level becomes independent of transistor characteristics. Bias resistors R81
and R82, with emitter resistor R83, establish the dc conditions in the
oscillator. The oscillator output is coupled through capacitor C40.
Power for the gray-frequency generator is supplied only when the noise
test mode has been selected by the TEST SELECT switch on the fm test
panel. (Refer to Table 5 -1 for the frequency obtained for each operating
standard. )
The black- frequency generator supplies the black- reference frequencies
used in the modulator afc module. It functions in a manner similar to that
of the gray generator just described, except that five oscillator output fre-
quencies can be obtained instead of three.
The white- frequency generator contains four switch-controlled crystals
that control the frequency of the white oscillator (Q5) .
Both the gray and black oscillators supply their outputs to transistor
switching circuits that permit only one oscillator output to be supplied at
any given time. Switch Q13 controls the gray -oscillator output, and switch
Q12 controls the black-oscillator output. Either switch is de- energized
according to the position of relay K3, which biases off either Q12 or Q13.
The unbiased output is then amplified through fm- driver subassembly Al
and supplied to the modulator afc module as the black or gray fm reference
frequency.
For the white insert and timing description to follow, see Fig. 5 -26.
For components discussed, see Fig. 5 -24. The white oscillator and crystal
switches remain inoperative until they are energized through relay Kl.
With reception of a white -insert signal from the fm filter module, K1
operates and connects operating power to the white oscillator and its asso-
ciated switches.
The white -reference frequency is injected into the fm path by an elec-

tronic switching arrangement using diode bridges X1 /X2 and X3 /X4.
Normally, the bridges are biased so that fm is supplied through the elec-
tronic switch. However, when white insert is selected, the bias is reversed
for short intervals, and the white -reference frequency instead of fm is
supplied during those intervals. The timing intervals of the reference
switch and fm switch are such that the reference switch permits 37-micro-
second pulses to pass during a 250 -microsecond interval of vertical blank-
ing (Fig. 5 -26).
A white - insert command from the fm filter module actuates relay K1,
which energizes the white oscillator and enables diode gate X33. Simul-
taneously, K1 allows the 6H pulses supplied from the tape sync processor
module to trigger the 250- microsecond interval -gate multivibrator (Q48,
First Horiz Sync After Vert Interval h63.5 ps -I
Sync
(For Time Ref) -4.8 ps
6H Pulse
P1 -29
Interval
Q48 -C
Back
Gate
-Porch
7
Pulses 1lJt -bus
P1 -27
r-
Delay MV --i F-12 Vs
Q46 -C
White -Insert
MV 37 ps-
Q51 -C
White -Ref
X1/X2
Output
FM
X3/X4
Output
FM + Ref
P1 -31
Demodulated
White Pulse From
Std Ref Fre4 /
Video
Video Signal 37 ps-

U u U u u
--I --12 us
Fig. 5 -26. Timing diagram, white -insert interval.

Q49) As shown in Fig. 5 -26, the positive-going trailing edge of the 6H

.
pulse triggers the interval -gate multivibrator, which provides a negative

level of 250- microseconds duration to the diode gate (X39, X40) .
Diodes X39 and X40 form a negative -input AND gate. The diode gate
remains enabled for the 250 -microsecond interval of the negative level.
During this time, a series of back-porch, or trailing-edge (TE), sync pulses
supplied from the tape sync processor module and passing through gate
X33 have been continuously triggering the 12- microsecond delay multivi-
brator (Q46, Q47) This multivibrator effectively delays the trailing edge
.
of sync by 12 microseconds and is constantly supplying delayed trigger

pulses to the diode gate at a horizontal rate.
The diode gate is enabled for 250 microseconds; therefore, only a
limited number of delayed sync TE pulses pass through the gate. Those
that pass trigger the white -insert multivibrator (Q50, Q51) , which sup-
plies an equal number of 37- microsecond pulses. These pulses are split in
phase through Q52, and the opposite polarities are amplified through Q54
and Q55 and applied at opposite ends of the quad diodes in bridge X1 /X2.
This is the white -insert diode bridge that functions as the reference switch.
The white -oscillator output frequency is supplied from Q5 to the input
of quad diodes X1 /X2. As the 37-microsecond pulses are supplied to the
quad diodes, they forward bias the diodes and permit the white -oscillator
output to pass through the reference switch. Thus, during vertical inter-
vals, white frequencies are supplied to the input of feedback amplifier
Q44 and Q45.
The 37-microsecond pulses supplied from the white -insert multivibrator
are supplied to two phase splitters simultaneously. One output is fed to
phase splitter Q52 (described previously) , while the second output is fed
to Q53. The opposite polarities from Q53 are amplified through Q56 and
Q57, and these outputs are applied at opposite ends of quad diode bridge
X3 /X4. This is the fm diode bridge and functions as the fm switch.
Tape fm is supplied from the fm equalizer module during playback.
In the E -E mode, E -E fm is supplied from the record equalizer module,
and relay K2 is activated by the E-E control signal from the modulator afc
module. Thus, in either mode, fm signals are supplied to the input of the
fm switch.
The polarities of the white -insert gating signals supplied from the 37-
microsecond multivibrator are arranged on both diode bridges so that one
bridge is forward biased while the opposite bridge is simultaneously reverse
biased. Thus, the fm diode bridge is forward biased except during the time
of white insert. This permits fm to pass at all times except for 37-micro-
second periods immediately following the vertical -sync interval, at which
times the white -insert diode bridge becomes forward biased and allows the
white frequency to pass.
The fm signal with the white frequency inserted is amplified through
feedback amplifier Q44, Q45. Unity-gain amplification takes place in the
output amplifier (Q58, Q59) to supply an output of 0.5 volt peak -to -peak
to the fm filter module.
Negative -going reference vertical pulses are supplied from the linelock
module to the discharge switch (Q40) The latter produces ramp sweep
.
signals that are used during video -head resonance tests. The sweep level
is set by a front -panel control, which sets the current level from current
source Q41 to charge the ramp capacitor (C61) .
The ramp pulses formed across the capacitor are coupled to comple-
mentary emitter followers Q42 and Q43. The pulses are then supplied to
the modulator module where they substitute for video signals from the
video input module during head tests. The linear fm sweep provides re-
sponse data which are used during the head antiresonance adjustments.
5 -6. COMBINED MODULATION- DEMODULATION

FREQUENCY RESPONSE
Basic to understanding the video -fm function in tape recorders is a
knowledge of the respective response characteristics of each portion of the
signal path. Current standards specify that all pre-emphasis and de- emphasis
be placed in the video portion. This then requires that the rf portion of the
signal path (fm carrier) have flat response over the passband of interest.
In practice, this means that the record current in the video heads (actually
fm) must be independent of frequency over the passband of interest.
Fig. 5 -27A shows the effect of head resonance and the correction neces-
sary in the rf playback portion to compensate for this effect. Due to head
resonance and various characteristics of the recording path, the record
equalizer (Fig. 5 -27B) contains critical adjustments normally made only
by the manufacturer at the factory. The overall result must be indepen-
dent of head record currents with respect to frequency in the passband.
This obviously requires very tight specifications between different video -
head assemblies.
Any variations in head response that do occur are compensated for in
the rf portion of the playback system by means of individual head "equali-
zation" controls (Fig. 5 -27C) Shown in Fig. 5 -27C is the antiresonance
.
network necessary to compensate for the head characteristic of Fig. 5 -27A.

The variable resistor is a Q, or damping, control to adjust the peak ampli-
tude of correction (also shown in Fig. 5 -27A) .
After compensation, the fm carrier (rf response) should be "flat" going

into the demodulator, as in Fig. 5 -27D. After video is extracted from the
rf, the video de- emphasis circuit complements the video pre- emphasis
applied ahead of the modulator (normally in the video input module) .
Video pre- emphasis and de- emphasis standards are set by the SMPTE
de- emphasis curve of Fig. 5 -28A. The pre- emphasis curve is complementary
to the de- emphasis curve-simply turn the curve "upside down" and change
the dB values from minus to plus to obtain the pre- emphasis curve.
Playback -Amp
Head Resonance Curve
Response Curve
FM in From Mod
4 Rec
Contains Channels
\ Pre -Emphasized
Video
to Heads
\.......--
¡
Adj
"Crossover"
Freq r
Peak Adj High Med Low
ti
10 16
Frequency IMHz)
(A) Head resonance (B) Record equalization

and compensation. arrangement.
-i
From
Head Other
Channels
To
Demod
6 dB/ Individual
Antiresonant
Octave Head
Circuit
(C) Head -response compensation.
Slope Equalizer
Head
Linear Phase vs Amplitude
De- Emphasis
From Playback Amp Demodulator "Flat" Video
(Video)
via "Flat" FM
(D) Demodulator input and output.

Fig. 5 -27. Frequency characteristics in signal path.
0 1.0
-2.0 0.8
_
8: -4,4 á 0.6
s HB
-8.0 0.4
LBM
-14.0 0.2
L BC
10 kHz 100 kHz 1 MHz 10 MHz

Frequency
(A) Response curves.
Standard RI : R2 TI us 12 us
Rl Low -Band 4:1 0.132 0.0264

Mono
Low -Band
6.56:1 0.24 0.03175
R2 Color
Eout High -Band 1.5:1 0.24

0.6
T1 ° IRI + R2IC T2 ° IR211C1
(B) De- emphasis network. (C) Resistors and time constants.

Fig. 5 -28. SMPTE de- emphasis.
Proper frequency and transient response of the system depend not only
on the magnitude of emphasis, but also on the shape of the curve.
Therefore, the slope of the curve is stated as a time constant.
The curves can be defined as the normalized impedance of the network
of Fig. 5 -28B. The SMPTE specifications can be simplified by putting
them in the more practical forms of Figs. 5 -28B and 5 -28C. The SMPTE
specifications are:
Low -Band Monochrome (LBM) :
T = 0.132 µs
X = 4.0
Low -Band Color (LBC) :
T=0.240,us
X=6.56
High -Band:
T=0.600µs
X = 1.5
where,
T is the time constant,
X is the ratio of R1 to R2 in Fig. 5 -28B.
This information is tabulated in Fig. 5 -28C.

In the de- emphasis network of Fig. 5 -28B, let:
RI = 320 ohms
R2 = 80 ohms
C= 330 pF
When C is in picofarads, put resistance in megohms to get the time con-
stant in microseconds. Then:
T1 = (R1 + R2) (C)
= (0.0004) (330) = 0.132 µs
T2 = (R2) (C)
_ (0.00008) (330) =0.0264µs
and
R1:R2=4:1
This is the proper de- emphasis network for low -band monochrome
operation.
Examination of the R2 -C network in Fig. 5 -28B shows that as the fre-
quency increases, the reactance of C becomes less, and the total impedance
therefore becomes less. For example, on the high -band curve of Fig. 5 -28A,
C becomes a dead short at 0.4 relative amplitude ( -8 dB) at 4 MHz.
In practice, the actual values of the resistors are modified by the imped-
ances coupled back. For example, refer to the pre- emphasis (opposite
curve to de- emphasis) networks of Fig. 5 -29. These show the kind of actual
values you will find in such networks in the latest systems. In these circuits
the tolerance is 1/4 percent for resistors and I/2 percent for capacitors.
Let us digress briefly here to give you the proper perspective. The
video and fm system has been designed to very exacting specifications. The
net result is that third and fourth generation color dubs can be made that
will provide quite suitable reproduction. Obviously, this requires excellent
frequency and transient response, and low noise level, in the master record-
ing. If you do not make dubs, you can get by with emergency parts replace-
ments that are not exact. But it should be emphasized that you should
obtain exact replacements as soon as possible. It all depends on how far
you are willing to let the overall performance slip over a period of time.
(Chapter 12 treats maintenance procedures.)
Note that in the networks of Fig. 5 -29, capacitor C shunts R1 so that as
the frequency increases, the net amplitude to the common -base amplifier
increases. Exactly the same time constants are used as in the de-emphasis
network which complements the particular mode of operation.
But, as stated earlier, the actual values of components must be modified
for the type of circuit used. If you calculate the time constants in Fig. 5 -29,
you will see that the period is somewhat longer than that listed in Fig.
5 -28C. But note how closely the ratio of R1 to R2 holds in each case.
207.9 pFR 138.3

R2 R2 PF RI
Emitter Common- Emitter Common -
Follower Follower 264.5 1735 Base Amp
158.7 634.8 Base Amp
(A) Low -band monochrome. (B) Low -band color.
1110 1471 pF
(C) High band. Emitter
R2
Common -
Follower 407.8 Base Amp
271.9
Fig. 5 -29. Examples of actual pre- emphasis networks.
We have now progressed in our studies to the record equalizer function.

Review Fig. 5 -1 and note that this circuitry drives four record amplifiers
(one for each individual head on the headwheel). These amplifiers provide
a constant rf current across the passband so that frequency variations
without amplitude changes occur. Approximately 100 volts peak -to -peak
of fm carrier exists at the headwheel brushes in the recording mode of
operation.
NOTE: Early tape systems employed electronic quadrature correction cir-
cuitry in the record or playback mode of operation, or both. Modern
video heads are constructed to within plus or minus 5 nanoseconds of
quadrature error, and no electronic quadrature correction circuitry is used.
5 -7. THE VIDEO PREAMPLIFIER

We will now look into the playback function of the system. In the play-
back mode of operation, only about 5 millivolts of signal exists at the input
to the video preamplifier.
Fundamentals of the video preamplifiers for the Ampex Mark X head

assembly are illustrated in Fig. 5 -30. Note from Fig. 5 -30A that during
the play mode the cathode of the nuvistor is biased through a resistor from
approximately -11 volts. This allows the nuvistor to amplify the repro-
duced signal which is coupled to its grid. The signal at the plate of the
nuvistor is routed through a low- capacitance coaxial cable to the associated
individual playback amplifier. This amplifier in turn supplies the individ-
ual-channel playback amplifiers containing the head -compensation circuitry.
During the record mode, the nuvistor is cut off because its cathode is
biased by +50 volts through RK. The nuvistor cathode impedance and CK
Record Voltage
From
Record Amplifier
Low -
Capacitance Playback Amp
Coax r
Nuvistor Input
Rotary 20
Type 7895
Head Transformer
220
Damping
í2 Resistor Degeneration
mH
ill( Required) Resistor
(If Required)
Mark X Video Head -J
Assembly and Nuvistor
Preamplifier
L +84 V
-I1 V Play
+ 50 V Record o
(A) Nuvistor preamplifier.

Phase and
Low -
Frequency
T fide Cl RI
Compensation
Amplifier IIJ, E16
LI -R 25 05, 06
Reproduced RF i
From Nuvistor Feedback Feedback
Preamplifier on o-e. Amplifier
Amp ifier
09
Amplifier TP1
Mark X Video Q10, 011 07,08
Head Assembly
(Factory -Set
C7 To Jl
Response) RF Output
(B) Overall block diagram.

Fig. 5 -30. Video playback amplifier.
form a time constant. Capacitor C,, is necessarily small to permit the

nuvistor to turn off rapidly enough to match the switching requirements
of the Editec accessory ( Chapter 10 ) The relatively small capacitance pre-
.
sented by C1 causes a phase error in the low rf frequencies during the play
mode. However, this error is compensated in the individual playback
amplifiers.
The video playback amplifier and the video record amplifier are physi-
cally placed on the same etched board, but they are completely separated
electrically and are powered by separate power supplies. There are four
such etched boards (one for each channel); each is located in one of the
four head -channel units. Fig. 5 -30B identifies the circuit functions and
relationships of the playback preamplifier. Each of the playback amplifiers
receives the rf signal (after amplification in an individual one -stage
nuvistor preamplifier) reproduced by the video head with which it is
associated. Each of the four nuvistor preamplifiers is integral with the
Mark X head assembly.
The output of the associated nuvistor preamplifier enters the playback
amplifier at pin 2 by way of a low- capacitance coaxial cable of specific
length. The signal at pin 2 is applied to the emitter of Q11, which presents
a very low input impedance. Transistors Q10 and Q11 form a feedback
amplifier. The combination of the nuvistor and transistors forms a low -
noise cascode circuit, of which the nuvistor is the input section and feed-
back pair Q10 -Q11 is the output section.
The output of Q10 and Q11 is amplified by Q9 which produces a gain
of approximately 2. Components Ll and R25 reduce the current feedback
of Q9 at low frequencies, thus increasing the gain produced at these fre-
quencies. This action compensates for the low- frequency losses and phase
shift in the cathode circuit of the nuvistor stage. Between 1 and 15 MHz,
Ll and R25 provide phase compensation; at frequencies below 1 MHz,
they provide increasing gain as the frequency is lowered. Their principal
purpose is that of phase compensation to maintain phase linearity of the
reproduced video signal.
The output of amplifier Q9 is applied to the base of Q8, which with Q7
forms a feedback amplifier. This stage produces a gain of approximately 6.
Variable capacitor C7 is factory adjusted to eliminate gain variations of
fractions of a decibel at 15 MHz (arising from variations in transistor
parameters or tolerances of other components). The combination of the
nuvistor preamplifier, interconnecting cable, and playback amplifier is
factory set by means of C7 to be flat within less than 0.1 dB between 1 and
15 MHz. (This is typical of many later stages in the video signal path.)
NOTE: The chain composed of the nuvistor preamplifier, the inter-

connecting cable, and the playback amplifier is arranged to produce a
gain of 50 (± 10 percent) when terminated in 75 ohms, and when there
is no unbypassed resistor present in the cathode circuit of the nuvistor.
This means that a 250 -millivolt output from the playback amplifier re-
sults from an output of approximately 5 millivolts from the head trans-

former.
The output of Q7 and Q8 is applied to the base of Q6, which with Q5

forms a complementary totem -pole amplifier ( two-stage amplifier with a
feedback factor of 1). The input and output voltages of this stage are the
same, but the output impedance is of the order of 1 ohm.
5 -8. THE PLAYBACK AMPLIFIER
Each playback amplifier (one in each channel, driven by the output of

its associated preamplifier as just described) performs the following
functions:
1. Compensation of head resonance.
2. Linear phase -amplituderesponse correction (cosine equalization of
frequency response).
3. Automatic gain control (agc) .
4. Fm level sensing for monitoring.
The circuit of each video head includes its own characteristic inductance,
resistance, and stray capacitance plus the input capacitance of the associated
nuvistor preamplifier. This causes resonance within the bandpass of the
video head. The condition of resonance causes a change of group delay
which must be compensated. These effects of head resonance are removed
by the playback amplifier, the characteristics of which are made exactly
opposite to those of the head.
Fig. 5 -31 is a block diagram of the playback amplifier in the RCA TR-
70 equipment. To accomplish its functions, three major signal -flow paths
are used. Each of these is described separately below.
Normal Signal Flow

During normal playback operation, relays K1, K2, and K4 remain de-
energized, and playback fm from the associated channel in the fm pre-
amplifier module is supplied to emitter follower Q6. This transistor drives
two paths as follows:
1. The primary path, through slope- amplifier transistor Q5, which pro-
vides head -resonance compensation, linear phase -amplitude response
correction, and agc.
2. The secondary path, through emitter follower Q7, which provides fm
level detection for display on the cro.
In the primary signal path, transistor Q5 amplifies the playback fm signal

in such a way that the output voltage increases with frequency at approxi-
mately 6 dB per octave. Slope amplifier Q5 drives the succeeding stages
containing transistors Q1 through Q4 and the head- resonance compensa-
2.2.15,4 ,74 PLO

.41
-§¡
E
ZgL
Courtesy RCA
Fig. 5-31. Block diagram of RCA TR-70 playback amplifier.
tion controls in the emitter circuit of Q3. These controls ( REACTANCE

and RESISTANCE ) are located on the front panel.
Complementary characteristics provided by the head-resonance corn-
pensation circuit offset head phase nonlinearities in the head-resonance
region. In addition to these complementary head characteristics, the head -
resonance compensation circuit introduces a rolloff of approximately 6 dB
per octave, which is offset by slope amplifier Q5.
The output of the head -resonance compensation circuit is taken from
the collector of Q1 and fed to delay -line driver Q31 -Q32. The latter feeds
the signal through a line -matching potentiometer to a delay line and
emitter follower Q33.
The delay-line output drives emitter follower Q35, which operates with
feedback amplifier Q36-Q37 as a difference amplifier. The output of
emitter follower Q33 drives resistor R85, which operates with diode shunt
X1 to form a variable voltage divider. The voltage at the junction of resistor
R85 and diode shunt X1 is applied to Q36 in the difference amplifier.
A combination of the functions of the delay line, variable voltage divider,
and difference amplifier provides cosine equalization of the playback fm
signal. Cosine equalization is required to compensate for frequency-
response rolloff that occurs as a result of the finite gap width in the heads.
Equalization corrects this rolloff of amplitude response while maintaining
a linear phase response. The amount of correction is determined by the
setting of the EQUALIZER control on the module front panel.
The equalized output from transistor Q37 in the difference amplifier is
fed to another variable voltage divider consisting of resistor R105 and
diode shunt X2. Voltage at the junction of R105 and X2 is amplified by
feedback -pair amplifiers Q38 -Q39 and Q40 -Q41. This voltage divider has
its voltage level determined by the agc feedback loop (described below) ,
which controls the impedance of diode shunt X2.
After amplification through the feedback -pair amplifiers, the signal is
fed to series amplifier Q42 -Q43, from which the signal is routed to two
paths. These are:
1.To the center arm of relay K4, then to either pin 31 as the playback
fm output, or to pin 30 as the head-resonance test output.
2. To emitter follower Q44 and the agc loop.
A sending -end termination of 75 ohms is provided to the signal before

it is applied to the contacts of relay K4. A receiving -end termination of
75 ohms is provided at either the fm switcher module (522) or the fm
equalizer module (523) , depending on whether relay K4 is de- energized
or energized.
In the agc loop, the signal from series amplifier Q42 -Q43 is fed to
emitter follower Q44, amplified through Q15, and detected through Q14.
In level detector Q14, the fm signal is peak- detected to yield a dc voltage
that is proportional to the signal level. This dc level is fed through the
normally closed contacts of relay Kl to emitter follower Q13. The signal

is then coupled through emitter follower Q12 to Q11.
Transistor Q11 serves as a temperature compensator. This stage, with
emitter followers Q12 and Q13, drives diode driver Q10. As the signal
level increases, transistor Q10 is driven further into conduction, and there
is more current through Q10 and diode shunt X2; this reduces the im-
pedance of X2. The reduction of impedance reduces the signal level from
the variable voltage divider (resistor R105 and diode shunt X2), resulting
in a constant output level regardless of variations in the input signal level
to the module.
Signal Flow for RF Level Detection

In this signal path, the playback fm signal is applied through emitter
follower Q7 to amplifier Q8. The amplified signal is then supplied to level
detector Q9. The circuits for transistors Q8 and Q9 are similar to those for
transistors Q14 and Q15. However, the remaining portion of the peak -
detection circuit is located in the fm equalizer module, where all four
playback fm channels are combined into a single path.
Signal Flow for Head -Resonance Test
During head-resonance tests, an fm sweep signal is injected into one
head at a time while the associated playback channel is selected with the
CHANNEL SELECT switch on the fm test panel. In this test mode, relays KI,
K2, and K4 are energized, and the fm signal is applied only to the selected
playback amplifier module, which is modified as follows:
1. With relay K2 energized, the signal is inhibited from transistor Q33
but follows the delay-line path to the difference amplifier (Q35, Q36,
and Q37) , which now yields a flat response.
2. Energized relay K1 is now in the manual -gain -control position, and
the gain of the amplifier is controlled by the setting of the front -panel
MGC potentiometer.
3. Energized relay K4 now routes the signal to pin 30 and to the fm
equalizer module, where the fm sweep signal is amplified, filtered,
detected, chopped, and finally supplied to the cro for display of the
head- resonance presentation.
Two series regulators are employed in the playback amplifier module to
supply the minus and plus 12 volts (nominal) required in the module.
Transistor Q16 provides regulated minus 12 volts from the minus 20 -volt
bus; similarly, transistor Q17 provides plus 12 volts.
5 -9. ELECTRONIC CHANNEL SWITCHING
In the playback mode, the four video -head outputs must be sequentially
switched to minimize noise and cross talk and to prevent two heads from
reproducing identical information during the overlap period. Obviously,

switching must be timed with the head revolution. Also, some means is
provided to assure that any switching transients fall in the horizontal -
retrace period so that they will not be visible in the reproduced picture.
Both Ampex and RCA perform this function in two principal switching
operations. The first is a coarse 4 -to -2 switching action that combines the
outputs of heads separated by 180° on the headwheel; the second is a 2 -to-1
precise switching action that combines the resulting paired outputs into a
continuous signal. This method is illustrated in Fig. 5 -32. Note that the
only difference here is the head -numbering sequence of the two manufac-
turers (described in Chapter 2) In either case, the 4 -to -2 switching action
.
combines the outputs of heads that are 180° apart on the perimeter of the
headwheel.
Playback 4-to-2 Switcher 2-to-1 Switcher
Preamp Amp
00 00
Combined
Channels
00
corn
Paired
Channels
Combined
OOOO
--r
RF
anca Out
Combined
Ampex Numbering
Channels
O RCA Numbering
00
Ofl
Playback
Preamp Amp
00 OO RF Switcher
Fig. 5 -32. Principle of rf switching systems.
If the basic differences are kept in mind, a description of the RCA

TR -70 fm switching system will serve to outline the basic action of all
electronic switching systems used in quadruplex playback. This description
follows.
Purpose
The fm switcher module combines the fm outputs of the four heads as
they sequentially scan the tape during playback. Precision switching in two
separate channels (video and sync) places switching transients where they
are least detrimental. In the video fm channel, the switching arrangement

normally locates transients during the horizontal-sync interval, where the
transients are later eliminated by substitution of regenerated sync. In the
sync fm channel, switching occurs during the video portion of the playback
signal, thus leaving the sync undisturbed by timing errors and transients.
In addition, the sync fm channel has provisions for equalizing the fm prior
to demodulation. (The fm equalizer module performs this function more
accurately for the video fm channel.)
Besides combining the playback fm outputs from the four heads, the fm
switcher module also generates a head -switching pulse that is used to
activate a transient -suppression clamp circuit in the post- emphasis module
(529). The fm switcher also generates a delayed signal that is referenced
to the tonewheel output for use in the accessory Chrominance Amplitude
and Velocity Error Corrector (Cavec) module (626) (See Chapter 9.)
.
NOTE: The fm switcher module operates universally on all television

standards. The descriptions that follow are simplified by referring to
tonewheel rates only on the basis of a 60 -Hz system (240 -Hz basic
tonewheel frequency). For 50 -Hz systems, change the corresponding
tonewheel frequency references to 250 Hz.
Because of the complexity of signal flow in this module, the description

is divided into five groups. These groups are divided according to the func-
tions of the module. Associated diagrams for pertinent groups are provided.
4 X 2 Coarse Switch
See Figs. 5 -33 and 5 -34. This circuit receives four fm inputs and one con-
trol input. The four fm inputs are supplied from the playback amplifier
modules. Each of the inputs supplied to the fm switcher module is a short
period of fm having a duration that corresponds to the time that the asso-
ciated head is scanning a recorded tape. The periods of fm output from
each channel occur sequentially and overlap in time, since one head begins
its passage over the tape before the previous one has completed its scan.
A 240-Hz tonewheel (tw) pulse is used for the control input signal (see
waveform A of Fig. 5 -34). This tw pulse is generated in the tonewheel
processor module (612) Its timing is based on a signal generated in a
.
pickup coil by a notch in the tonewheel as the tonewheel rotates with the
headwheel. The tw pulse thus has a definite relationship with the rotating
quadrature heads. The tw pulse occurs near the center of the active tape
scan by head 1.
Two gates are used in the 4 X 2 coarse switch to combine alternate input
lines. Channels 1 and 3 are combined in the diode gate using X19 through
X22. (See Fig. 5 -34, waveforms B, D, and I.) Channels 2 and 4 are com-
bined in the diode gate using X8 through X11 (waveforms C, E, and J) .
Switching between input channels occurs in each gate during the interval
when neither input channel is supplying a tape fm signal.
.74
>
ax
1L gé
QO
ao
Fig. 5 -33. Block diagram

THE VIDEO AND FM SYSTEM 1 ;;
VI
Wo
of RCA fm switcher.
Switching between channels 2 and 4 is controlled by a monostable multi-

vibrator using transistors Q43 and Q44. When triggered by the tw pulse,
the multivibrator generates a 2080 -microsecond pulse (waveform F) that
corresponds to one-half the 240 -Hz tw period. The polarity of the control
signal from transistors Q43 and Q44 is such that the gate switches to chan-
nel 2 before head 2 begins to traverse the tape (switching actually occurs
at the tw leading edge while head 1 is at the center of the tape) After head .
2 has completed its traverse (2080 microseconds after the initial switch) ,
the gate switches from channel 2 to channel 4. Thus, disturbances in the
desired fm signal are avoided by gating the two input signals while both
are inactive.
A similar sequence of operation occurs in the gate for channels 1 and 3.
However, in this gate the tw pulse cannot be used directly to trigger the
FM
TW
240 Hz
Channel
FM
Input
Input
Channel 2
1
'
! siiiiii k
pr viriivii
Y
..Mni
rATA"
Input %G
Channel 3 &NM NM
FM Input
Channel 4 WAYA ' WA WAN
/i.
214 Control
.
2080µs
Q43 Out
i i
Delay Pulse
1040 µs
Q45 Out
fM
1/3 Control
Q46
FM 1/3
Out
OI
-
Mn.
-
\
Gate Out
v, v
%'/iii,y °
M3 & M9 In
FM 2/4
Gate Out (
---
M2 &M9In
FM Video
%/l
Ch Sw Out
M2/M3 Out
M2 Anode
M3 Cathode®L
480 Hz
-
OK /X///i %//3iii7 %/.iiR'//
Fig. 5 -34. Timing diagram for 4 X 2 coarse switch and video channel.
control multivibrator (Q46 and Q47) This would cause switching to occur
.
during the active head traverse because the tw pulse is generated near the
center of the active scan for head 1. To overcome this problem, the tw
pulse is delayed for a time interval corresponding to 90° (1040 micro-
seconds) of rotation of the headwheel by delay -generator transistor Q45
(waveform G in Fig. 5 -34) The trailing edge of the delayed output pulse
.
from Q45 then triggers the delayed 240 -Hz multivibrator (Q46 and Q47,
waveform H) and causes the gate for channels 1 and 3 to switch between
the intervals when heads 1 and 3 are supplying fm.
Output from the gate for channels 1 and 3 ( waveform I) is fed through
buffer transistors Q21 and Q20 to one input of the video -channel fm switch
and one input of the sync -channel fm switch, respectively. Similarly, the
output from the gate for channels 2 and 4 (waveform J) is fed through
buffer transistors Q9 and Q8 to the second inputs of the video -channel and
sync -channel fm switches.
The delay generator (Q45) provides an auxiliary output for the fm
switcher module through pin 10 to the Cavec module ( 626) when the
latter accessory is used with the machine.
Video -Channel FM Switch
Two diode quads constitute an electronic switch controlled by precisely
timed signals developed in the timing -generator circuits of the fm switcher
module. One input to the electronic switch is the combined fm from chan-
nels 1 and 3 at quad diodes M3 (waveform I of Fig. 5 -34) The second
.
input is the combined fm from channels 2 and 4 at quad diodes M2

(waveform J) .
The outputs of the two quads are connected in parallel to drive subse-
quent stages. The control points of the quads are connected in complemen-
tary fashion to the push -pull control signals so that the quads select the
inputs alternately as a single -pole, double -throw switch. Switching occurs
at each polarity reversal of the control signal, so the switch operates at a
960 -Hz (4tw) rate when supplied with a 480 -Hz control signal (wave-
form L) .
The 960 -Hz switching rate causes the quad diodes to select the channel
1 or 3 input while head 1 is traversing the tape. The quad diodes then
switch to the channel 2 or 4 input, just as head 1 is leaving the tape and
head 2 is beginning its scan. As head 2 completes its scan, the quad diodes
switch again to the channel 1 or 3 input to select the output of head 3 as it
starts across the tape. When head 3 is nearing the end of its active scan,
the quad diodes switch away and select the channel 2 or 4 input once again
as head 4 begins its traverse. Finally, as head 4 nears completion of its scan,
the quad diodes return to the channel 1 or 3 input to begin the sequence
again with head 1.
The output from the quad diodes is a continuous fm signal (waveform
K in Fig. 5 -34) resulting from the sequential selection of playback chan-
nels coordinated with the passage of the heads across the tape. Because the
headwheel rotates 240 times per second, complete cycles in the selection of
all four heads must occur at a 240 -Hz rate, and selection of individual heads
occurs at a 960 -Hz rate.
Video -channel fm in its entirety (waveform K) is fed to a diode switch
(X4 and X5) , which is controlled by the record bus. This switch is open
for operation in the record mode to prevent high -level recording fm, which
may be picked up in the sensitive fm preamplifier module, from proceeding
through the fm switcher module. Thus the switch eliminates possible cross
talk in later stages.
Fm through the switch is supplied to a feedback line driver (Q6 and
Q7) that restores the signal level lost in the switching circuits and the
output termination. A high -pass filter is used between the line driver and
the output terminal (pin 17) to allow usable fm to continue to the fm
equalizer module (523) , but to reject low- frequency offsets caused by
minor dc unbalance in the 2 X 4 gates and M2 -M3 switches.
Sync -Channel FM Switch and Equalizer
Refer to Figs. 5 -33 and 5 -34. Diode quads M9 and M10 in the sync
channel function in the same manner as M2 and M3 in the video channel,
except that timing of the switching process is advanced in the sync chan-
nel. However, unlike the arrangement in the video channel, the sync-
channel fm does not pass through a separate amplitude -equalizer module
prior to demodulation. Instead, playback losses are compensated for by
simplified equalizers with fixed adjustments in the fm switcher module.
Two separate equalizers in the module are optimized individually, one for
high band and the other for low band. Relay Kl selects the proper equalizer.
It is energized by standards bus 4 or 5 for 525- or 625 -line high -band
standards, and it remains de- energized for all low -band standards.
Either equalizer operates in conjunction with common -base amplifier
Q40 to boost high- frequency fm signal components relative to the low -
frequency components. Sync- channel video is used only to provide a source
of tape sync undisturbed by switching pulses; therefore, adjustment of the
equalizers is not so critical as in the video channel, and fixed controls can
be used.
Equalized fm is fed through buffer emitter follower Q39 to the feedback
line driver (Q37 and Q38) . The fm is then filtered through a high -pass
filter as in the video channel.
Timing Generator
The complexity of the timing -generator signal flow requires the descrip-
tion to be in three parts, each having a seperate block diagram and timing
diagram.
Video Channel, Normal Operation-For this description, refer to Fig.
5 -33, the simplified diagram in Fig. 5 -35, and the waveforms of Fig. 5 -36.
Sync-
Inverter
Channel G30
Switch
4TW - !
960-Hz
Bistable MV
C141, 042
EF
Q16
12yu
Pulse Gen
Q15
Video - InputGate
Channel (AND)
Switch SI Q14
l
480-Hz To Video -Channel
TN/ Bistable MV Switch Driver
Q!, Q2
Fig. 5 -35. Simplified diagram of video -channel timing generator

for normal operation.
A sync -channel switch pulse having a repetition rate of 15,750 Hz is gen-

erated in the internal reference module (Chapter 6) and enters the fm
switcher module through pin 26 (see waveform A of Fig. 5 -36) The .
pulse is inverted by transistor Q30 and triggers the 960 -Hz bistable multi -
vibrator (Q41 and Q42) An additional trigger pulse for the multivibrator
.
is the 4tw signal (waveform B) which is developed in the tonewheel

processor module (523) and enters through pin 28. The tonewheel
processor module contains the 4tw delay control that permits adjusting the
4tw pulse to the proper point in the head overlap interval.
The bistable multivibrator (Q41, Q42) is switched to one state by a
positive -going transition of the 4tw signal that is coincident with the
beginning of a new head traverse across the tape. The next sync -channel
switch pulse to occur then causes the bistable multivibrator to change state
again. This results in a negative pulse from transistor Q42 with its leading
edge timed by the 4tw pulse and its trailing edge timed by the sync -channel
switch pulse (see waveform C in Fig. 5 -36) The multivibrator is immune
.
to triggering by further switch pulses and is inhibited until it is again reset

by the 4tw trigger as the next head begins crossing the tape.
Negative pulses from transistor Q42 are supplied at a 960 -Hz rate to
emitter follower Q16, which drives pulse generator Q15. The pulse from
Q15 (waveform D in Fig. 5 -36) has a duration of approximately 12
microseconds, with its leading edge timed by the leading edge of the sync -
channel switch pulse.
The 12- microsecond pulse from Q15 is fed to one input of the input gate
(Q14). A second signal to the input gate is the video- channel switch pulse
(waveform E, Fig. 5 -36), which is generated in the internal reference
module. This second pulse has a repetition rate of 15,750 Hz and a nega-
2 ps Division
Sync Channel
Switch Pulse
4TW
1-
r
Adjustable With 41W De ay
960 -HV MV
Q42 Output
Pulse Gen 12 us r
Q15 Output
Video Channel
Switch Pulse O
I put Gate
n
Q14 Output I
FM
N.
M2 Input
M3
FM
Input & N. \, 1
Head Overlap'
%. 3
FM
M21M3 Output
. \
FM Interval Corresponding
Q2 Output
O to Horizontal Sync
M2 Cathode
M3 Anode O ¡(
M2 Anode
M3 Cathode O I
1040 ys/Division
FM 214
Gate Output if0 /AM /AM 'MW OEM %/í.L/
M2 Input
FM 113
Gate Output
H NIKV
/
M3 Input
FM
M2 /M3 Output
I
4 WASFA MM VA 1
Q2 Output
480 Hz
M2 Cathode (
M3 Anode
480 Hz
M2 Anode
M3 Cathode
O
240 HzTW
'Actual overlap is several hundred microseconds.
Fig. 5 -36. Timing diagram for video channel, normal operation.

tive -going transition timed during the horizontal -sync interval at the
precise instant that video- channel head switching is to take place. This
signal is connected to the second input of the input gate when switch S1
is in the Normal position.
The input gate produces a positive output when both inputs are negative.
This condition exists during the time interval between the leading edge of
the 15,750 -Hz video- channel switch pulse and the end of the 12- micro-
second 960 -Hz pulse (see waveform F, Fig. 5 -36) . In summary, the char-
acteristics of this video-channel trigger pulse are determined as follows:
1. Leading edge: Video- channel switching pulse
2. Trailing edge: Sync -channel switch pulse plus a 12- microsecond
interval generated in transistor Q15
3. Repetition rate: 4tw (960 Hz)
A positive -going output from transistor Q14 triggers bistable multivi-
brator Q1 -Q2, which operates as a two -to -one frequency divider. The
square-wave output from Q1 -Q2 is at a 480 -Hz rate (waveform J, Fig.
5 -36) and is clipped by shunt limiter M4 (Fig. 5 -33). The signal is then
supplied to the video- channel switch driver, where push -pull control signals
(waveforms K and L, Fig. 5 -36) are developed to drive the video- channel
precision switch (M2 and M3 in Fig. 5 -33) The latter supplies a continu-
.
ous fm signal ( waveform I) from the inputs shown in waveforms G and H

of Fig. 5 -36. Proper phasing of the multivibrator is assured by introducing
a tw pulse ( waveform M) to reset it in the event that the phase is in-
correct.
NOTE: Head overlap of adjacent channels is several hundred micro-
seconds, but it is shown greatly reduced in the timing diagrams to simplify
illustration of the various functions.
Video Channel, RF Copy Operation -For this description, refer to Figs.

5
4TWo -
-33, 5 -37, and 5 -38. Playing back or recording a tape using the rf -copy
960 -Hz
Bistable MV
041, Q42
.a EF
016
12 µs
Pulse Gen
015
n pu Gate
Inverter I
(AND)
Q30 o 014
SI
Sync -
Channel o
Switch
To
Horiz -Gate 480-Hz Video -Channel
I nverter Pulse Gen
Delay Bi stable MV Switch
019 017
018 01,02 Driver
TW o
Fig. 5 -37. Simplified diagram of video -channel timing generator, rf -copy mode.
mode of operation ( direct re-recording of fm without intervening demodu-

lation to video and subsequent modulation) requires the timing of the
video -channel switching point to be changed. During normal video play-
back, head switching is positioned in the sync interval. This results in
elimination of the switching transient from the video output by subsequent
replacement of tape sync with regenerated sync. However, during playback
of an rf copy recording, sync cannot be replaced because the signal was not
demodulated at the time the copy was made. Therefore, switching is posi-
tioned in the front -porch interval, and switching transients are clamped
out by the post- emphasis module (529) during final video playback. The
advance in switching time is accomplished by replacing the video -channel
switch pulse at the input gate (Q14) with another pulse having an
adjustable delay; this occurs when switch S1 is placed in the rf copy
position.
2 Ns/Division
Video Channel
Switch Pulse O
Sync Channel
Switch Pulse
Inverter
Q19 Output
Adj by
Sw Delay-1
Horiz
r
Gate Delay OD l'
QI8 Output
Pulse Gen
Q17 Output
E
-- - - - - -- fi -1/ _---- - -i_ -J
Pulse Gen ( 7)
Output O 12 ps
Input
014 Output
Gate
G
s.. -a. -1
- - - -I` -_ - -_.
FM
M2 Input
Head Overlap*
O
¡
FM
M3 Input
Exact Position of Sw tch Depends
on Switch Delay Setting
FM
M21M3 Output 3
J
FM Interval Corresponding Position of Switch During Sync
to Front Porch Interval in Normal Operation
Fig. 5 -38. Timing diagram for video channel, rf -copy mode.

A new, adjustable delay pulse is derived from the sync -channel switch
pulse (waveform B of Fig. 5 -38) , which occurs almost six microseconds
in advance of the video-channel switch pulse ( waveform A) . The sync -
channel switch pulse is inverted by transistor Q19. The resulting output
pulse drives horizontal-gate delay generator Q18. Here, a new delay pulse
is generated; its leading edge is coincident with the input pulse, but its
trailing -edge transition time ( waveform D, Fig. 5 -38) is determined by
the position of sw DELAY potentiometer R85.
The variable -width pulse from transistor Q18 drives pulse generator Q17,
where a new constant -width pulse (waveform E) is formed. The leading
edge of this pulse is coincident with the trailing edge of the input pulse.
Thus, the position of the pulse is determined by the setting of the Sw
DELAY control.
Pulse generator Q17 supplies the variable-delay pulse at a 15,750 -Hz
rate. The pulse is combined with the 960 -Hz, 12- microsecond pulse (wave-
form F) from pulse generator Q15 in input gate Q14. (The pulse from
Q15 is generated exactly as in normal operation.) The input gate produces
a positive pulse (waveform G) each time the inputs from Q15 and Q17
are both negative.
Sync- Inverter
Channel ----jl 030
Switch
960-Hz
4TW Bi stable MV -a EF
Q33
Q41,042
1W
480-Hz
Bistable MV
Q31, Q32
-- ToSync-Chan nel
Switch Driver
Fig. 5 -39. Simplified diagram of timing generator for sync channel.
The resulting pulse from the input gate has a positive -going leading
edge that triggers the 480 -Hz bistable multivibrator (described previously) .
However, the multivibrator output transitions occur during the front -porch
interval, and quad diodes M2 and M3 are switched accordingly. Thus, the
output of the quad diodes (waveform J) is a continuous fm signal repre-
senting sequential switching between heads, with the switching point
occurring in the front -porch interval as adjusted with the SW DELAY
control.
Sync Channel-Refer to Figs. 5 -33, 5 -39, and 5 -40 for this description.
The 4tw input pulse (waveform B, Fig. 5 -40) causes the 960 -Hz bistable
multivibrator to change state as a head is just beginning to traverse the
tape. After the multivibrator has changed state, the next sync -channel
switch pulse to occur passes through inverter Q30 and returns the multi-
vibrator to its original state, where it remains until it is reset by the next
4tw positive trigger. This results in a 960 -Hz negative pulse from the
multivibrator (waveform C, Fig. 5 -40). The leading edge of this pulse is
determined by the positive -going 4tw pulse transition, and the trailing
edge is timed by the sync -channel switching pulse.
Sync-Chan (7C\
Pulseo
4TW
960-Hz MV
Q42 Output
480-Hz MV
Q32 Output
M9 Cathode
E
M10 Anode
M9 Anode
t
MIO Cathode I
Input
OM9 FM : Head Overlap
nput
M10 IFM
H
FM
M9/M10
Output
OI ZARIBBarA
Picture Interval
\L\
Horiz- Blanking Interval

Fig. 5 -40. Timing diagram for sync channel.
Up to this point, timing generation is identical to that in the video-

channel timing generator for the normal mode (see waveforms A, B, and
C of Fig. 5 -36.) However, in the sync -channel timing generator, instead
of being applied to additional gating circuitry, the pulse from the 960 -Hz
multivibrator (Q41, Q42) is immediately coupled through emitter follower
Q33 to trigger the 480 -Hz bistable multivibrator (Q31, Q32). In addition,
the tonewheel pulse is fed to the same 480 -Hz multivibrator to assure
proper phasing of the output.
Output 480 -Hz pulses from transistor Q32 (waveform D in Fig. 5 -40)
are clipped by quad diodes M5 (Fig. 5 -33) and then coupled to the sync-
channel switch driver (transistors Q34, Q35, and Q36) Here, the push -pull .
control signals (waveforms E and F) are developed to drive the sync-

channel precision switch consisting of quad diodes M9 and M10. The

precision switch combines the outputs from the gate for channels 2 and 4
and the gate for channels 1 and 3 into a continuous fm signal (wave-
form I) The switching point in the signal is now before horizontal
.
blanking.
Head -Switch Pulse Generator
Gate Q10 (Fig. 5 -33) is enabled by the record bus whenever the
machine is in any mode other than record. At those times, the gate passes
480 -Hz pulses from the 480 -Hz multivibrator (Q1, Q2). Each transition
of the 480 -Hz signal triggers the monostable multivibrator (Q11, Q12),
and the multivibrator generates seven -microsecond pulses at a 960 -Hz
rate. This is now the head- switching pulse output, the leading edge of
which coincides with the video-channel switch point. During input or
noise test conditions, the post- emphasis module utilizes the head -switch
pulse to clamp out switching disturbances in the demodulated video signal.
5 -10. MODERN DEMODULATION CIRCUITRY

Demodulation circuitry, since it is now strictly solid- state, has changed
considerably from that described earlier in this chapter. But, if you under-
stand pulse-width demodulation (described earlier) , you also understand
the basic action of modern demodulators. Essentially, this process depends
on the fact that as the carrier frequency increases, the pulses grow closer
together, and as the carrier frequency decreases, the pulses grow farther
apart. A frequency change therefore results in a dc change (dc pulsing) ,
which a discriminator can convert to video information.
After switching and equalization, the signal -to-noise ratio is improved in
the fm filter module. This module reduces noise without affecting the signal
to be demodulated by redistributing the sideband information and by
attenuating noise. Balance in the sidebands is restored in the limiter stages
of the viedo-channel demodulator which follows.
In addition to restoring the sideband balance, the limiters in the video -
channel demodulator module also serve to remove amplitude variations
caused by variations in head -to -tape contact and differences in tape content
level (e.g. scratches or similar disturbances) . After substantial limiting,
the signal is supplied to a discriminator that converts frequency variations
in the fm signal to a varying -amplitude signal.
Demodulated video supplied from the video- channel demodulator con-
tains carrier and modulation components that must be removed in order to
return the signal to characteristics similar to those of the record signal prior
to modulation. The video filter module removes these undesired com-
ponents and supplies a low -level video signal to the post- emphasis module.
The video-channel signal is amplified and clamped in the post- emphasis
module. The high -frequency boost (pre- emphasis) that had been added in
the video input module is now removed, and the video is restored to its
original form. A separate section in the module is used to process the sync -
channel signal similarly, so that both channels maintain separate character-
istics as their signals are supplied to the video processing system ( next
chapter) and the monitors.
Sync-channel demodulation (Fig. 5 -1) is performed with the sync-
channel demodulator module (used by RCA) , which is identical in con-
struction to the video -channel demodulator module. Using a separate
channel for sync demodulation provides a desirable enhancement of picture
quality and sync stability. This is accomplished by having sync-channel
switching occur during fm overlay corresponding to the picture interval.
Thus, timing errors are not introduced, and switching disturbances do not
occur during the demodulated sync interval. Conversely, the video-channel
switching is accomplished during sync -interval fm overlap, leaving the
active picture interval undisturbed. The disturbance to the sync pulse
caused by switching is overcome by completely replacing the sync pulse
with a clean pulse derived form sync demodulation in the processing
amplifier (Chapter 6) .
NOTE: In the Ampex system, the processing amplifier derives and pro-
vides switching pulses which cause video head switching to take place
on the front porch of sync, thus avoiding any ambiguity in the timing
edge of horizontal sync.
See Fig. 5 -41. In the video -channel demodulator, the signal is split in
phase and processed in a push -pull manner through the module. Limiter
stages clip the signal to remove variations in level, and push -pull fm is
supplied to the discriminator stage.
Conversion of frequency variations to amplitude variations is initiated
in the discriminator circuit where constant -width pulses are generated at
a repetition rate corresponding to the instantaneous carrier frequency. The
effective video level from the discriminator is very low compared to the
carrier level. However, the filter following the discriminator partially re-
moves those carrier components that may cause cross talk, and it supplies
the push-pull video to an amplifier stage to provide the module output.
Push-pull output from the demodulator module is supplied to the video
filter module, where selection is made for the proper Bode -type filter
according to the operating standard of the machine. The Bode filters present
a characteristic that passes video and attenuates fm. The constant-width
pulses from the demodulator discriminator are integrated, forming video
that passes relatively undisturbed through the passband region; however,
in the stopband region, the phase is caused to lag by 90° in one half of the
filter and lead by 90° in the opposite half of the filter. This lag and lead
condition occurs at higher frequencies, which is where fm cancellation is
required.
+12 V
Cascode
Diff Amp
From Phase To Next

% DIN Amp
Splitter
+12 V
(A) Partial simplified schematic.
Phase
Splitter
Diode Detector
1I
Cascode
5 Similar
Differential Diff Amp
Stages
Amp
J
To Filtering
8 Output
Limiter 1
& Emitter
Follower
ll
Diff Amp
(B) Block diagram.

Fig. 5 -41. Basics of solid -state demodulator circuitry.
Let us now study Fig. 5 -41 in more detail. Fm supplied to either

the video -channel demodulator or sync -channel demodulator is fed to a
phase splitter, from which outputs of opposite phase are supplied to a dual -
channel cascode differential amplifier.
The cascode differential amplifier supplies its outputs to the first of five
limiter stages. Each stage operates in push -pull with an associated differen-
tial amplifier and emitter follower for each channel. The opposite phases
of the fm signal are balanced, amplified, and symmetrically clipped prior
to combination of the phases for demodulation.
Output from the fourth limiter stage is supplied to a differential ampli-
fier as well as to the following fifth limiter stage. This differential amplifier
then feeds an output from the module through a relay (not shown) when
the rf -copy mode of operation is selected. This output, which is still in fm
form, is supplied from the machine as the rf -copy signal and is sometimes
used directly for rerecording when a copy tape is to be made.
The output of the fifth limiter stage is coupled through emitter followers
to the delay -line drivers. Reflections from delay lines DL1 and DL2 pro-
duce narrow pulses as a result of the square-wave current pulses that enter
the delay lines (Fig. 5 -42) Each half of the push -pull circuit generates a
.
positive or negative pulse at each corresponding input square -wave transi-

tion. The pulse repetition rate is therefore twice the instantaneous fm
carrier frequency.
The quad diodes in M1 (Fig. 5 -41B) are connected as a bridge rectifier
that gates the push -pull positive and negative pulses from the delay lines to
the proper inputs of the filter drivers. The bridge supplies only negative
pulses to point 2 and only positive pulses to point 1.
Each delay line (DL1 and DL2) consists of a precisely cut 84 -inch
length of 75 -ohm foam coaxial cable that is shorted at its outer end. The
input network of each line provides an accurate impedance match through
the frequency band of 1 MHz to above 100 MHz. This is necessary be-
cause the pulses present include frequency components in this frequency
range. The first reflection of each square -wave current pulse that enters
the delay line is a 17- nanosecond pulse; the second and third reflections
are cancelled out. (The second reflection is less than 4 percent of the first;
the third reflection is less than 2 percent of the second.)
With time and amplitude balance of the 17- nanosecond positive and
negative pulses established, the only remaining variable is frequency, which
is directly related to the interval between pulses. The action of the bridge
detector doubles the number of positive and negative pulses that are
present. The frequency at which there is no interval between pulses occurs
at approximately
1
2(17 x 10 -9sec) -29MHz

The voltage at the output of the bridge detector rises linearly with propor-
THE VIDEO AND FM SYSTEM
Fig. 5 -42. Simplified schematic of delay -line drivers and diode detector.
tional increases of frequency from zero to the maximum at approximately

29 MHz.
Integration (through filtering) of the constant -width pulses from the
bridge rectifier develops a voltage with an amplitude that is dependent on
only the pulse repetition rate. Thus, in the push -pull phase with positive
pulses, the higher the pulse repetition rate becomes, the more positive will
be the output voltage. A similar situation exists in the negative phase of
the push -pull circuit, where a negative output voltage is developed.
NOTE: This chapter has not discussed the recovered video signals follow-
ing demodulation. The remainder of the video system includes special
processing of the signal components, and is more appropriately studied
in Chapter 6.
EXERCISES
Q5 -1. Under what circumstances could electronic "quadrature correctors" be
eliminated in both the record and playback functions?
Q5 -2. What is the fm bandwidth of (A) older tape systems and (B) later -
model combination high -band, low -band systems?
Q5 -3. Give the corresponding frequency for each of the following:
(A) Low -band sync -tip level (monochrome)
(B) Low -band blanking level (monochrome)
(C) Low -band peak white (monochrome)
(D) Low -band color sync -tip level
(E) Low -band color blanking level
(F) Low -band color peak white
(G) High -band sync -tip level
(H) High -band blanking level
(I) High -band peak white
Q5 -4. Define "dropouts."
Q5 -5. Approximately how much rf voltage exists at the headwheel brushes
in (A) the record mode and (B) the playback mode?
Q5 -6. What is the primary signal that supplies switching information to the
electronic channel switcher?
Q5 -7. Where does the input signal to the demodulator come from in (A) the
record mode and (B) the playback mode?
Q5 -8. Why is fm amplitude limiting used in the demodulator?
Q5 -9. Why must timing information from separated sync be fed to the
electronic switcher?
Q5 -10. Is head -resonance compensation used for fm equalization?
CHAPTER 6
The Video Processing System
Fig. 6 -1 presents basic block diagrams of the circuitry to be studied in

this chapter. In Fig. 6 -1A, MATC means Monochrome Automatic Time
Correction, and CATC means Color Automatic Time Correction. Fig. 6 -1B
is the Ampex equivalent of the upper row in Fig. 6 -1A, indicating differ-
ences in arrangement and nomenclature from RCA.
The video paths of Pig. 6 -1 are being observed on the system output
monitoring equipment during the E -E or record mode of operation. During
the play mode, this video path is being continuously corrected by various
error signals to be described. The connected groups of blocks in Fig. 6 -lA
are interconnected in various ways by internal harnessing. External lines in
and out of the individual blocks are indicated on the drawing.
6 -1. FUNCTIONAL DESCRIPTION

The basic terminology for and functions of the blocks shown in Fig.
6 -lA are as follows:
Tape Sync Processor: This unit separates tape sync from the sync -channel
video and provides optimum noise immunity and transient gating for all
sync functions.
MATC Error Detector: The time -base error of the separated tape sync
is measured and converted to the necessary control signal for MATC.
MATC Video: This section contains the variable delay line and driver
circuits for the MATC function.
CATC Video: This section contains the variable delay line and driver
circuits for the Color ATC function.
Color Error Detector: Color time -base errors are detected to provide
the control signal for Color ATC.
Burst Processor: This unit provides separation of the burst from the tape
signal for the color error detector. It also includes circuits for shaping the
new burst from the local subcarrier.
151
Fig. 6 -1. Video processing systems.

Color Phase: This circuit provides adjustment of regenerated burst

phase and system subcarrier phase.
Nonphased Color: This section provides necessary pulse circuits for
control of the nonphased color mode ( sometimes termed NIC, or noninter-
laced color) .
Internal Reference: This part of the system provides afc locked to the
tape horizontal sync to provide precise timing of all switching, transient
suppression, sync gating, and ATC pulses.
Vertical Advance: Special circuitry counts the number of pulses in a
field to determine accurately the position for regenerated vertical blanking.
Sync Logic: This unit generates horizontal and vertical blanking and
combines them into composite blanking. It combines tape sync and regen-
erated horizontal sync into composite regenerated sync, and it generates a
start pulse which phases the counting of the vertical advance circuitry.
Horizontal AFC: Sync separated from the color- corrected video signal is
used to control the frequency and phase of a multivibrator. This, in com-
bination with other circuits, generates new horizontal sync, front porch,
and blanking.
6 -2. BASIC PRINCIPLES OF ATC
The Ampex terminology is Time Error Compensation (Amtec) for

monochrome and Color Time Error Correction (Colortec) for color. The
RCA terminology is Automatic Time Correction (ATC or MATC) for
monochrome and Color Automatic Time Correction (Color ATC or CATC)
for color.
One action of ATC circuitry is to control the speed and phase of the
headwheel, in addition to the control exerted by the servo system. There-
fore, the servo systems are mentioned in this study, and Table 6 -1 spells
out the differences in terminology used by Ampex and RCA for the
specified modes of servo operation. (Servo systems are covered in detail in
Chapter 7.)
The operation of ATC is based on application of an error signal to an
electronically variable delay line inserted between the video output of the
demodulator section and the input of the signal -processing amplifier. The
error signal is obtained by comparing the timing of horizontal sync in the
composite video playback signal at the demodulator output with the timing
of reference sync. Since geometric distortion is an error in time, it can be
corrected by a variable delay line of opposite time function.
The ATC circuits operate in either of two modes, internal or external.
The internal mode is used only when the headwheel servo is in non -
synchronous operation, and the external mode is used only when the head -
wheel servo is in synchronous operation. The circuits sense whether the
system is in tonewheel (Ampex normal) or Pixlock (Ampex automatic)
operation and automatically switch to the internal or external mode.
In the internal mode, ATC corrects geometric distortion but does not
synchronize vertical- or horizontal -sync pulses from the tape recorder with
the corresponding sync pulses from the local sync generator. To accomplish
this type of correction, the reference sync is timed by deriving an average
from a long- time -constant afc loop locked to tape sync. A dc error signal
is obtained by making a phase comparison between a sample pulse
derived from the trailing edge of tape sync and a trapezoid waveform
generated from reference sync. The trapezoid timing is steady enough to
eliminate geometric distortion but changes sufficiently with the average
timing of tape sync to prevent slow drifts from causing the error signal to
exceed the correction range of the delay line.
Table 6 -1. Servo Terminology
Mode
No. Ampex RCA Servo Mode
1. Normal Tonewheel Tachometer signal referenced to local 60

Hz, but no feedback to servos from demod-
ulated tape sync. Nonsynchronous mode of
operation.
2. Vertical Switchlock Same as mode 1, but phase comparison of

local 60 Hz with demodulated tape 60 Hz
"frames" tape playback with local signal.
Provides "no- roll" switching, or vertical -
synchronous operating mode.
3. Horizontal Linelock Phase comparison of local horizontal rate

and demodulated tape horizontal rate locks
tape playback to local horizontal rate. Not
necessarily "framed" but synchronous (in
time) to local horizontal.
4. Automatic Pixlock Phase comparison of demodulated tape sig-

nal horizontal and vertical rates to local
horizontal and vertical rates provides "gen-
locked" operation of tape system. Fully syn-
chronous (time and phase) to allow lap -dis-
solves, supers, special effects, etc.
Note 1: Operating mode 3 or 4 is required or color tape playback.
Note 2: The Ampex AVR -1 video output is always fully synchronous in any mode of operation.
Time -base error is corrected so that the off tape video signal will be in sync with the system
input reference signals. At the output, reference sync, blanking, and burst signals are added to
the time- corrected video signal. Thus the recorder -output composite video signal is synchronous
(except for delay of cabling outputs) with all other video signals that are connected to the
same sync generator.
In the external mode, essentially the same circuits are used, but the afc
loop is locked to local sync instead of tape sync. Since the local sync signal
is fixed in timing, the error signal derived by comparing it with tape sync
contains information to correct both jitter and geometric distortion. As
a result, the ATC greatly reduces residual Pixlock (automatic) jitter and
provides an extremely stable output.
In either of the two modes, the error signal is amplified and applied to
the delay line. As the demodulated composite video signal passes through
the delay line, the length of the line is modulated by the error signal every
63.5 microseconds, or once per tv line.
Full correction range is maintained automatically by a tha ( tape horizon-
tal alignment) feedback circuit. This circuit adjusts the relative phasing be-
tween the ATC sample pulse and the ATC trapezoid so that the average
error signal is in the center of the delay -line range. In the internal mode,
the tha signal controls a variable delay in the internal reference afc loop.
In the external mode, the tha signal causes the Pixlock servo to advance
the phasing of tape sync with respect to local sync by controlling the head-
wheel motor speed.
Geometric distortion includes various segmentation errors which result
in picture distortions such as jogs, scallops, and quadrature -error indenta-
tions (Chapter 2) Segmentation errors are sufficiently minimized for
.
monochrome picture reproduction by the monochrome ATC system; how-

ever, optimum color reproduction requires a considerably greater error
reduction. A comparison of segmentation error effects between mono-
chrome and color reproduction may be made by observing the fact that a
picture displacement from line to line of 0.03 microsecond (30 nano-
seconds) is not visible in a monochrome picture but represents a 38°
color -subcarrier phase shift, which in turn produces a serious hue shift in
the color picture.
Time-base jitter, inherent in all quadruplex recording processes, is
caused by nonuniformities in the mechanical motions of the machine.
When the machine is in the tightest mode of servo operation ( Pixlock or
Linelock for RCA, automatic or horizontal for Ampex) the servo system
,
is capable of reducing the residual jitter to a value of approximately 0.1

microsecond, which is entirely satisfactory for monochrome operation but
which represents a 125° color -subcarrier phase shift and is, therefore, intol-
erable for color operation. (The foregoing statement is true because 0.8
nanosecond of error represents one degree at the color -subcarrier frequency
of 3.58 MHz, as derived in Chapter 3. Since 0.1 µs = 100 ns, then
100/0.8 = 125 °.)
ATC completely corrects steps (such as the errors resulting from im-
proper head quadrature) because the time error remains constant through-
out each individual line. The same is true for jogs (skewing) because the
timing error from jogs ( improper tip penetration) although increasing
,
throughout the line, increases at a constant rate and is the same for cor-
responding points on successive lines.
Errors resulting from scalloping cannot be removed completely by an
ATC system, because the timing error at the right-hand side of the picture
is not the same for successive lines. The difference in timing errors for
successive lines will be seen in the ATC output as a small skew -like effect
in vertical lines at the right -hand side of the picture only. It is, therefore,
essential that the manual scalloping adjustment on the headwheel panel be

performed carefully.
Practical field experience with quadruplex television tape recorders has
shown that it is reasonably easy to maintain the total time -base error below
one microsecond peak to peak. Therefore, an automatic timing corrector
must have a correction range of about one microsecond to permit non-
critical operation of the recorder. This range is achieved by the use of
electronically variable delay lines.
NOTE: It is imperative that the reader be familiar with two basic circuit
functions very common in ATC circuitry, or any circuitry which involves
error detection. These are (1) boxcar pulse generators (delayed and un-
delayed) and (2) sampling -pulse (sample and hold) circuitry. These are
covered in Chapters 9 and 12, respectively, of Workshop in Solid State
by Harold E. Ennes, (Indianapolis: Howard W. Sams & Co., Inc., 1970).
In addition, the application of voltage -variable- capacitance diodes in volt-
age- controlled delay lines must be understood. This is covered in Chapter
2 of the same reference.
Time -base error correction is inserted between the demodulated video

output and the input to the video processing amplifiers (Fig. 6 -2) Assume
.
the errors of Fig. 6 -2A. (This is the condition of Fig. 2 -14, Chapter 2.)
Waveform 1 is the sync time recorded, and waveforms 2 and 3 show the
playback sync timing for the first and last lines of a complete band of 16
lines. Waveform 4 shows that this results in a series of linearly progress-
ing errors across each band of 16 lines, or the duration of one head pass.
Now go to Fig. 6 -2B. Since errors must be sampled (measured) before
a correction is made, a fixed delay of 2 to 3 µs is employed, as shown, prior
to application of the corrector voltage to the voltage -controlled video delay
line.
Reference sync changes with the mode of operation. For nonsynchronous
operation, this reference is separated horizontal sync from the video signal
after demodulation. For synchronous operation of the servos, the reference
becomes local horizontal sync from the station sync generator. The termi-
nology here is "internal" for nonsynchronous modes and "external" for
synchronous modes.
In the internal mode, tape sync supplies both the sampling pulse and the
trapezoid. The trapezoid is obtained by applying tape sync to a long -time-
constant afc loop. This mode provides a relatively large range of control,
since the trapezoid changes sufficiently with average timing of tape sync
to prevent drifts from exceeding the characteristic impedance (Z0) of the
variable delay line.
When the voltage -controlled variable delay line is modulated away from
its center of operation, Z. becomes increasingly mismatched (mistermi-
nated) When a given range of correction (delay) is exceeded, reflections
.
from mistermination occur and upset the servos. The maximum of all errors
in the time base should not exceed 1 µs.
Recorded Q -u Amplitude of Error
1st line of band of

16 lines enters latee AAA
Last lined band of
16 lines leaves early.' I
U¡
i
(A) Time-base error.

Corrected Video
Demod Fixed abl
To Proc Amps
Out Delay De ay
Uncorrected Video
Controls
Tape Sync Snc
Phase in Sep
Synchronous
Mode Fixed Ref Volts
Head Error
Motor Det
i 1
H ad Trap THA
Servo Gen Diff Amp
Delay Into, Variable

Control Ext Delay
Local Horiz Ref
(B) Correction circuitry.
In Out lot
Dellay
Delay /
L
Sample
i
+ Ext
Ref
Error Volts
(C) Variable delay. (D) Error waveforms.

Fig. 6 -2. Time -base error correction.
In the external mode (synchronous mode) , station sync is used, and

the head servo, in turn, is locked to this sync. This naturally imposes a
tighter restriction on the amplitude of tolerable errors from the head-to-
tape contact, as well as all other errors in servo and signal processing.
When color time -base correction is used, the station color standard
becomes the reference for that unit. Old burst is gated out and a new
burst, referenced to the stable subcarrier standard, is gated on. The error
between the old burst and the reference color standard is fed back to the
monochrome time -correction differential amplifier. Hence, in turn, the
head servo (Chapter 7) is tightened to its maximum capability. It is now
being controlled to correct to the color -subcarrier time base. This total
correction will result in well under 5° of jitter at the color -subcarrier rate
of 3.58 MHz.
6 -3. MONOCHROME ATC
Since a large timing shift may occur when the recorder switches from
one head to another during playback, it is essential that timing errors be
measured after switching so that the proper amount of corrective delay is
provided for the next line. In older systems, switching took place during the
horizontal -sync pulse, and the trailing edge of tape sync was used for
deriving the timing -error signal. In the latest systems, switching takes
place at a predetermined time ahead of the leading edge of sync, and the
leading edge of tape sync is used for deriving the timing-error signal.
Operation of the electronically variable delay line depends on the fact
that the capacitance of a reverse -biased silicon diode (varicap diode)
changes with applied bias. A variable delay line could be constructed
simply by connecting a number of inductors in series, connecting one
diode from each junction of the coils to a common bus, and applying a
variable dc voltage between one end of the series inductors and the com-
mon bus. However, this design would not be satisfactory because the
varying bias used to change the line length would appear in the output
signal. To overcome this, a second set of diodes is used, and each diode of
this set is connected from one of the junction points to a second bus (Fig.
6 -3) A fixed negative dc voltage is applied to the center (inductive) por-
.
tion of the line. A variable positive voltage is applied to the common bus
of one set of diodes, and a variable negative voltage is applied to the com-
mon bus of the other set. The bus supplying the positive voltage is called
the positive error bus (PEB) , and the bus supplying the negative voltage
is the negative error bus (NEB) The dc levels of the PEB and NEB
.
signals are fixed to equal amounts above and below the voltage applied to
the center, and ac components are equal but oppositely phased. The video
input is applied between one end of the coil series and ground, and the
delayed video output appears between the other end of the series and
ground.
THE VIDEO PROCESSING SYSTEM I?`)
Ma
It
.:.
r:.
"--{ 1
I
Fig. 6 -3. Electronically variable

delay line.
ECM
IF
KM"
L:.
r:
t:
mElt.
The electronically variable delay line used in a monochrome ATC system

consists of two halves connected in series. The PEB and NEB signals are
applied to the first half, and a pair of error signals delayed 1.5 microseconds
behind the first pair are applied to the second half. The delayed signals are
called the delayed positive error bus (DPEB) and delayed negative error
bus (DNEB) signals.
The following description involves both Fig. 6 -2 and the timing diagram
of Fig. 6 -4 in explaining the basics of monochrome ATC. Tape sync from
the demodulator output (Fig. 6 -4, waveform 2) is obtained by use of a
Microseconds
o 1 2 3 4 5 6 8
I I I I
Re(
Horiz Sync
(System
01 I
Sync Input)
Demodulated
Horiz Sync O r
ATC
Sample Pulse
(From Tape
Horiz Sync)
ATC
Trapezoid
(From Local
Horiz Sync)
Fig. 6 -4. Basic MATC timing diagram.
clamped sync separator designed to prevent introduction of any timing

errors by the separation process. Sync is then applied to a circuit which
generates a sampling pulse, approximately 2 microseconds wide, from the
leading edge of tape sync. The leading edge of this pulse (waveform 3)
coincides with the leading edge of sync. The sample pulse is applied to an
error detector which compares the timing, or phase, of the sample pulse
with a reference trapezoid waveform (waveform 4) In the internal mode
.
of ATC, the trapezoid is obtained by applying tape sync to a long-time

constant afc loop; in the external mode, local sync is used instead of tape
sync. Note also (Fig. 6 -2) that in the external mode the headwheel servo
system is controlled.
The error detector provides an output voltage which changes linearly
with the difference in phase between the sample pulse and the trapezoid.
Since the delay line has a nonlinear relationship between the voltage ap-
plied and the time delay, a linear signal cannot be used to drive the delay
line directly. To overcome this difficulty, the error signal is applied to an
amplifier whose output varies nonlinearly with the input signal (Fig.
6 -5A). The amplifier is carefully designed to match the delay -line character-
R2
Error In Nonlinear *2.2 V

-1.75 V
(A) Circuit arrangement.

3.
I.--"Maximum Delay
3. 4 '
3. 3
3. 2
3. 1
Zero Error and Nominal Delay Point

3. 0
2. 9 I
I
I
I
2. 8 I t
I I
2. 7
I ¡ Minimum Delay
2. 6
2.0 S
05 10 15 2.0 25 30 35 40
PE BI NE B Volts
PEB & NEB 0;25 1.1
Volts
I I 1 I I I I I
NLA Out 0.5 1.0 2.2 3.0 4.0 5.0 6.0 7.0
Volts
I I I
NLA In -1.08 -1.30 -1.75 -1.95 -2.15 -2.45 -2.6

Volts
(B) Delay versus voltages.

Fig. 6 -5. Delay -line drive in RCA TR-70 tape system.
istic so that the overall relationship is linear. The output of the nonlinear
amplifier is applied to a phase splitter which provides the PEB and NEB
signals, and through a fixed delay line to another phase splitter which pro-
duces the DPEB and DNEB signals. Each of the four error signals is
applied through an individual complementary-symmetry emitter -follower
stage to the variable delay line. Circuits of this type are used because they
provide low output source impedance on both positive -going and negative-
going signals to the error buses.
To recapitulate, the picture -channel video from the demodulator is fed
to the monochrome ATC video module. The video is amplified, clamped,
and applied to an electronically variable delay line. The video signal is
either delayed or advanced from the nominal error delay of 3.05 micro-
seconds ± 0.05 microsecond (Fig. 6 -5B) The delay line is controlled by
.
the positive error bus and negative error bus (PEB and NEB) signals,
which are developed from a phase comparison of the tape sync and the
reference sync in the ATC error detector module. The output of the MATC
video module is fed to the Color ATC video module where further error
correction takes place. Essentially, this is the most important function of
the MATC system. All of the other circuits in the system only contribute
to the detection of an error between tape sync and reference sync and the
controlling of these auxiliary circuits.
When the tape recorder is operating in the tonewheel servo mode, the
tha error signal is fed to the ATC trapezoid circuit in the ATC error
detector module, where it shifts the phase of the ATC trapezoid waveform
with respect to the ATC sample pulse. The tha loop is stabilized when the
PEB voltage is equal to the tha reference voltage of +1.1 volts. Under
these conditions, the dc average of the error signal keeps the electronically
variable delay line in the center of its range. It should be noted that this
condition is necessary to obtain maximum correction range (Fig. 6 -5) .
When the tape recorder is operating in the synchronous servo mode,

the tha error signal is fed to the Linelock module, where it causes the head -
wheel motor to shift the phase of the video signal from the tape. In turn,
this results in equal shifts in phase of tape sync and the ATC sample pulse
with respect to the ATC trapezoid waveform. The resulting change in the
PEB voltage increases or decreases the tha error signal. As in the tonewheel
mode, the loop is stabilized when the PEB level is equal to +1.1 volts.
6 -4. COLOR ATC
The preceding paragraphs contain a discussion of basic considerations

for an ATC system which operates effectively with monochrome signals.
For color operation, additional considerations must be taken into account.
One of these is noise. When horizontal sync is used in video -signal timing
measurements, only one sample per line is provided, and a relatively wide
bandwidth is required to pass the signal; therefore, timing measurements
are affected by random noise which is added by the tape recording process.
This effect is known as positional noise. Positional noise occurs at or below
the threshold of visibility during monochrome operation and is not critical;
however, during color operation, positional noise represents a substantial
phase jitter which results in phase instability. Color ATC systems reduce
the effect of positional noise because of the availability of a greater num-
ber of samples per line due to the presence of approximately eight cycles
of subcarrier in the burst, and because of the possible use of a narrower
bandwidth.
A second factor o'hich prohibits the use of horizontal -sync phase com-
parison in color operation is that there is no specified phase relationship
between the edge of sync and the color subcarrier. In addition, it is possible
that time modulatior may be introduced between horizontal sync and the
color subcarrier; therefore, horizontal sync cannot be used in controlling
the absolute phase of a subcarrier. It is essential that the final color timing
correction be accomplished by comparing the phase of burst with a fixed
reference.
The MATC open -loop system is capable of maintaining an error- reduc-
tion factor of at least 25 to 1, which is sufficient for monochrome reproduc-
tion but may leave enough residual error to disrupt color performance if
large errors are to be corrected. Because of the cascading effect of the
CATC process, residual error theoretically can be reduced by another fac-
tor of 25 to 1, and any error remaining is then negligible, even for color
operation.
Color ATC systems employ the same open -loop control described above
for monochrome signals; however, in color systems, the error detector uses
burst instead of horizontal sync. Also, CATC systems require a delay range
of only 360° at the subcarrier frequency (0.28 microsecond at a subcarrier
frequency of 3.58 Mhz); therefore, the delay line is shorter than that
utilized in the MATC systems and need not be split to contain the delay -
change transient within the horizontal -sync interval. The CATC nonlinear
amplifier, fixed delay line, and driver circuits are identical to those of
MATC systems.
In CATC systems, it is necessary that the phase -error detector have the
equivalent of a linear delay range of 360° at the subcarrier frequency.
Design limitations prevent the detector from functioning effectively over
an entire 360° delay range; therefore, the detector is operated at half-
frequency, and the required 360° delay range appears to the detector as a
180° range. Multivibrators divide both the burst and reference subcarrier
signals; the divided burst is used to form narrow sampling pulses, and the
reference signals form a sawtooth waveform. A special inhibit circuit in
the burst path insures that the narrow sampling pulses always occur on the
center half of the sawtooth slope.
The control signal for CATC variable delay lines is formed in the color
phase- detector circuit by the measurement of burst phase error. The
function of the control signal is to correct phase errors in composite color
signals and stabilize signal timing in order to provide excellent color repro-
duction. However, color signals must still be passed through processing -
amplifier circuitry to clean up blanking and sync and insert a new color
burst that is free of noise. This requires special circuits which do not
distort the color components of a signal.
The system used in processing color signals consists of splitting a com-

posite signal into high -pass and low -pass components. Low -pass components
do not contain color information and can be passed through standard
monochrome processing-amplifier circuits in order to clean up blanking
and sync. High -pass signals contain all chroma information and are
processed by special circuits in the CATC system to clean up the blanking
interval and insert regenerated burst. High -pass signals are then added back
to the low -pass signals just ahead of the output amplifier in the processing -
amplifier circuits.
Color ATC provides time -base corrections in tape playback signals to a
residual error level of a few nanoseconds with respect to the subcarrier
reference signal. This allows color tape recording with essentially no band-
width or color -response limitations; thus, a machine equipped with CATC
can record and play back color program material which is indistinguishable
from original live signals.
The color fixed delay line (Fig. 6 -6) inserts an additional 2.5 micro-
seconds of delay into the video signal path, and the output of the fixed delay
line is fed to the electronically variable delay line in the color delay module.
The variable delay line is controlled by the color PEB and NEB potentials,
which, in turn, are developed through a phase comparison of the tape burst
and reference subcarrier signals in the color error detector module (Fig.
6 -1) The phase-comparison process results in transients which appear on
.
the error buses. However, since tape burst information from which the
sample pulses are derived is obtained before the video signal has been
delayed by the color fixed delay line, transients applied to the variable
delay line advance 2.5 microseconds with respect to the video signal applied
to the line. Transients occur during the horizontal sync pulse interval and
do not affect burst or video information in a composite signal. The output
of the variable delay line is a composite video signal which is stabilized in
time with respect to the reference subcarrier. This signal is fed to the color
processing system.
Circuits of the color processor separate low- frequency and high- frequency
components of a delayed video signal so that they can be processed inde-
pendently before recombination. The separated high- frequency chroma
signal is fed to special processing circuits in the color processor module.
Special processing circuits for the high- frequency signal are required be-
cause the monochrome processing-amplifier circuits clamp out the burst
signal, and the black clipper removes all signal components below black
level. The low- frequency signal is fed to the MATC reference module
(Fig. 6 -6) Circuits in the MATC reference module separate sync from
.
the low -pass signal, and the separated sync is used to develop regenerated
sync in the sync logic module (Fig. 6 -1) Regenerated sync is fed to the
.
video output module where it is combined with the low -pass and high -pass
portions of the video signal (processed in the video control and color
processor modules, respectively) to form the video output signal.
Fig. 6 -6. Synchronous -mode servo with MATC and CATC.

As shown in Fig. 6 -1, the reference 3.58 -MHz subcarrier signal is fed to
the color phase module, which contains two electronically variable delay
lines. The first variable delay line is controlled by the SYSTEM PHASE
control and can be varied over a nominal range of 0.3 microsecond ( i.e.,
a delay equivalent to a subcarrier phase shift in excess of 360° ). Following
this delay, the reference subcarrier signal is fed simultaneously to the
second variable delay line in the color phase module and to the color error
detector module. The second variable delay line is controlled by the BURST
PHASE control, and the delay range available is also nominally 0.3 micro-
second. The reference subcarrier delayed by the second variable delay line
is designated the burst subcarrier and is fed to the color processor module.
The reference subcarrier fed directly to the color error detector module is
designated the system subcarrier.
Tape burst, separated from the video signal in the burst processor
module, is also fed to the color error detector module. In the color error
detector module, the system subcarrier signal is used in generating a refer-
ence sawtooth waveform; the tape burst signal is used in developing a
sample pulse. The sample pulse and sawtooth waveform are fed to a
phase-comparator circuit where the sample pulse falls higher or lower on
the sawtooth waveform slope, depending on the relative phase between
the reference -subcarrier and tape-burst signals. The output from the phase-
comparator circuit is an error signal, the magnitude of which reflects the
difference in phase between the system (reference) subcarrier and tape-
burst signals. Logic circuits insure that sampling occurs only over the
linear portion of the sawtooth waveform slope; therefore, the gain of the
error detector is linear. If the sample pulse falls near either extreme of the
range on the sawtooth waveform slope, it is possible to obtain a cracking
effect ( i.e., successive pulses sampling at alternate extremes of the slope,
which causes conflicting error information). To prevent this effect during
normal tape playback in the CATC mode, a portion of the phase -error
signal is used in controlling the thaf (tape horizontal alignment, fine)
signal in the CATC system; the thaf signal in turn adjusts the video delay
slightly so that the sample falls near the center of the sawtooth slope.
The burst logic circuit, which controls the sampling process, is in turn
controlled by the narrow burst -gating pulse developed in the burst
processor module. The width of this pulse is variable; however, the width
is nominally set at that value which allows approximately five samplings
during each line. The output error signal fed to the Color ATC module is
a fluctuating series of dc potentials. To eliminate the possibility of an
erratic output signal such as could occur if tape burst were suddenly lost,
a clamping circuit provides a steady dc output signal which corresponds
to zero error if the tape burst signal is absent. The clamping circuit is
controlled by the burst -lock sense potential which in turn is determined
by a logic circuit in the color processor module. The phase -error signal fed
to the CATC module is compared with a fixed reference in a differential
amplifier circuit, and the resulting output is developed into the PEB and
NEB potentials which control the variable delay line.
The PEB and NEB potentials which control the electronically variable
video delay line are developed from the phase -error signal of the color
error detector. The phase -error signal is an indication of the relationship
between the reference -subcarrier and tape -burst signals and varies linearly
with the phase difference between these signals. However, the character-
istics of the video delay line are such that the delay varies nonlinearly in
response to a linearly varying control potential impressed across it. It is
desired that the delay vary linearly; therefore, it is necessary to convert the
linearly varying phase -error signal to a nonlinear potential having gain
characteristics which match the delay characteristics of the video delay line.
The nonlinear conversion is accomplished by a differential amplifier circuit
that incorporates a gamma network containing resistors with values calcu-
lated to produce a nonlinear output which compensates for the nonlinearity
of the video delay line.
The nonlinear error signal is split in phase, and signals of opposite
polarity drive complementary- symmetry emitter followers which produce
PEB and NEB potentials. It is important that the PEB and NEB potentials
be equal in amplitude and opposite in polarity with respect to the voltage
to which the video signal in the delay line is referred, over the entire
range of the phase error signal. Therefore, potentiometers are provided as a
means of adjusting the PEB and NEB potentials to insure that a change in
one is accompanied by an equal and opposite change in the other.
Potentiometers are also provided as a means of adjusting the transient
behavior of the PEB and NEB potentials so that the rise times are as nearly
equal and opposite as possible in response to a step of error such as occurs
at switching when a vacuum -guide error is present.
Fig. 6 -6 shows the tha and thaf closed feedback loops to the headwheel
servo. The block diagram shows the functions of the headwheel servo,
MATC system, and CATC system that operate in conjunction with the tha
and thaf closed loops. In general, the principle function of the tha loop is
to adjust automatically the phase of the headwheel servo to maintain the
average sampling of the MATC in the center of the error detector slope.
This point, in turn, is set by internal adjustments to correspond to the PEB
potential for the center delay of the monochrome variable delay line.
The principal function of the thaf loop in the CATC system, during
playback in the normal CATC mode, is to adjust automatically the phase of
the MATC output signal to maintain the average sampling of the CATC
in the center of the error detector slope. The center of the error detector
slope in the thaf loop is also set up by internal adjustment to correspond
to the PEB voltage representing the center delay of the color variable delay
line. The actual operation is accomplished in several steps. First, the thaf
error signal is generated by comparing the PEB potential to a reference
voltage. The resulting error signal is applied, through a low -pass filter, to
the reference side of the input differential amplifier in the monochrome

nonlinear amplifier circuit. Therefore, a change in thaf error signal gen-
erates a change in the potentials of the MATC error buses, PEB, NEB,
DPEB, and DNEB, which are applied to the monochrome variable delay
line. The change in error -bus potential causes the delay of the mono-
chrome variable delay line to change in such a direction as to correct the
average phase of the signal applied to the CATC system.
In correcting the average phase of the signal applied to the CATC
system, the average delay of the monochrome variable delay line shifts
from the nominal center value of the PEB potential. The tha differential
amplifier in the MATC system measures this change in PEB potential and
generates its own error signal which feeds back to the phasing delay in the
Linelock module. A change in the Linelock phasing delay causes the head-
wheel to advance or retard so that the monochrome variable delay line is
restored to operation in the center of the delay range.
The net result of the closed -loop servo, during playback in the normal
CATC mode, is that the headwheel is changed in phase to supply the
correction demanded by the CATC system, the monochrome variable delay
line is still operating in the center of the delay range, but the MATC error
detector is slightly shifted from operation in the center of its slope. It
should be noted that the bandwidth of the thaf loop is somewhat greater
than that of the tha loop. Therefore, changes in the average thaf voltage
occurring at a frequency greater than the tha loop bandwidth are corrected
by changing the delay of the monochrome variable delay line rather than
by changing the phase of the headwheel.
There are several fixed delays shown in Fig. 6 -6 which are not basic to
the operation of the thaf loop during playback in the normal CATC mode.
The 7.5- microsecond tape horizontal delay and the 2-microsecond reference
horizontal delay are inserted into the tape horizontal and reference hori-
zontal paths, respectively. The net difference in the two delays represents a
5.5- microsecond delay in series with the Linelock phasing delay. The pur-
pose of this net delay is to maintain the operation of the variable Linelock
phasing delay at nominally the same value during playback in the normal
CATC mode as it is during playback in the MATC mode. At the same time,
this net delay increases the total delay to that value which advances the
headwheel by an amount equivalent to the delay inserted into the signal
path by the color fixed delay line and the color variable delay line.
NOTE: The arrangement of the fixed and variable delay lines in ATC
systems varies with the manufacturer and model of the equipment.
A fixed delay is also inserted into the path of the tape vertical signal
applied to the tva (tape vertical alignment) and lock-sense circuits of the
Linelock module. If the output signal from the machine is to be aligned
with reference sync as it normally is, the output from the demodulator
must be advanced considerably because of delays presented by the mono-
chrome fixed and variable delay lines and the color fixed and variable delay
lines. When the output is not in coincidence with reference vertical, the
vertical coincidence circuit will drop out and thereby throw the servo out
of the Pixlock mode. To prevent the machine from falling out of the Pix-
lock servo mode during color tape playback, tape vertical is timed to the
output signal by the fixed delay inserted into the tape vertical path.
In the nonphased color (npc) mode of tape playback, requirements are
somewhat different than in the normal CATC mode. The principal function
of the npc mode is to allow playback of noninterlaced color (nic) tapes.
Noninterlaced color tapes are produced by dubbing a playback from a
heterodyne color system. Inherent in the heterodyne color system is an
output signal in which the phases of the monochrome and chroma portions
of the signal are displaced by an amount equal to the jitter of the tonewheel
servo in the machine that made the original recording, with the result that
they no longer have a fixed relationship. The Linelock and the MATC
circuits, which are attempting to correct the video- signal time base by
comparing tape sync to reference sync, and the CATC system, which is
attempting to correct the video -signal time base by comparing tape burst
to reference subcarrier, are apparently presenting conflicting information.
However, the presentation of conflicting information is avoided by causing
the CATC system to assume control.
As the relative phase between the tape burst and reference subcarrier
attempts to change, the thaf differential amplifier senses this condition and
generates a signal which modulates a delay in the path of the reference sync
fed to the Linelock and MATC circuits. The headwheel servo and MATC
systems follow this change in phase of reference sync in such a manner that
the phase of the tape burst does not change with respect to the reference
subcarrier. The bandwidth of the thaf loop during color tape playback in
the npc mode must be greater than that of the tonewheel servo used in
recording the original tape. The net result is that the output signal from
the machine varies in phase in such a manner that the chroma signal bears
a fixed phase relationship to reference subcarrier. That is, the output signal
is noninterlaced color, but it is as satisfactory as the original noninterlaced
color playback.
To accomplish modulation of the reference horizontal delay with as great
a range as possible, it is necessary to provide a nominal delay of approxi-
mately one-half line. The nominal delay can then be modulated ± 20 micro-
seconds, if necessary, to follow the jitter of a very loose servo. A similar
fixed delay of approximately one -half line is inserted into the tape hori-
zontal path, so that the undelayed tape signal retains approximately the
same phase relationship to the undelayed reference sync during playback
in the npc mode that it does during playback in the normal CATC mode.
This allows the headwheel servo to operate in the Pixlock mode during
color tape playback in the npc mode, since vertical alignment is maintained.
Even though the tape signal is varying in time in response to modulation
of the reference horizontal delay, if its nominal position is centered with
respect to reference vertical, the excursions do not exceed the range of the
lock -sense circuit. In tape recorders equipped for the Linelock-only (LLO )
servo mode, it is advantageous to operate in that mode during npc playback
when vertical coincidence is not required. Since exact horizontal phasing is
not predictable in playback of a noninterlaced color tape (nor is it precisely
adjustable in playback of an interlaced color tape in the npc mode) , there
is no possibility of mixing, fading, or incorporating special effects, etc.
The prime purpose of the normal CATC mode is to provide these possi-
bilities during playback in the Pixlock servo mode.
Delayed reference horizontal, with the same modulation, must also be
fed to the MATC system. If the MATC system were operating from a fixed
horizontal reference, it would attempt to correct the timing modulation
which has been inserted into the tape signal by the headwheel servo. How-
ever, the delayed reference horizontal is delayed by nominally one-half line,
whereas the tape signal is nominally aligned to undelayed reference hori-
zontal. It is necessary to insert another half-line delay in the reference-
horizontal path to the MATC system to insure that the trapezoid waveform
is properly timed so that the sample pulse falls on its slope. This nominal
half -line delay is automatically adjusted to be precisely the correct amount
to cause average sampling at the center of the slope by the tha error signal.
It should be noted that the tha loop is relatively limited in bandwidth so
that the ATC is still performing a normal function of removing timing
errors such as quadrature error, jogs, etc. It should also be noted that the
thaf loop, during npc playback, is also bandwidth- limited so that it does not
respond to the normal line -by -line correction of phasing errors by the
CATC system.
There is one further refinement of MATC operation required for color
tape playback in the npc mode. This refinement consists of bypassing the
afc loop to form the ATC trapezoid directly from the delayed reference
horizontal. The afc loop performs no useful function in the external mode
of MATC operation; it is actually detrimental during color tape playback
in the npc mode, because it provides undesired bandwidth limiting and an
additional phase shift which makes the high -gain thaf loop very difficult
to stabilize. During color tape playback in the normal CATC mode, how-
ever, the afc loop provides a required timing advance which is not required
during npc operation because a delay of one full tv line is provided.
The remaining circuitry of Fig. 6-1 associated with the ATC system prior
to video processing -the tape sync processor and internal reference -will
be examined in the following two sections.
6 -5. THE TAPE SYNC PROCESSOR

The tape sync processor uses sync obtained during tape playback to
derive (1) afc tape horizontal sync, (2) ATC tape horizontal sync,
(3) vertical sync, (4) a tape frame pulse, (5) a back-porch clamp pulse,
'0
58
Q Ó
35
w Ó
Fig. 6 -7. Tope sync processor, RCA TR -70 system.

and (6) a 6H pulse. The tape sync processor module also derives horizontal
and vertical sync signals during back -to -back (E -E) operation.
As illustrated in Fig. 6 -7, tape horizontal sync is stripped from the
incoming video signal and is gated, at the horizontal rate of the system,
to two outputs by the leading -edge gate. One output is sent to the internal
reference module for use in the afc loop timing; the other is sent to the
ATC error detector module for use in MATC timing. Tape vertical sync
is stripped from the incoming video signal and is sent at the vertical (field)
rate of the system to the Linelock module, where it is used in the tva and
tm vertical circuits (Chapter 7) . The tape frame pulse is generated by
gating the tape horizontal and tape vertical sync pulses and sending the
resultant, at the frame rate (one-half the field rate) to (1) the capstan
phase module, where it is used in the capstan servo operation, and (2) the
splice logic module, where it is used for Switchlock detection (Chapter 10).
The back-porch clamp pulse is derived from the trailing edge of the
incoming composite-video sync signal and is sent, at the horizontal rate
of the system, to the MATC video module, where it is used to establish
the average picture level (APL) , and to the composite trailing- edge-gate
pulse generator to reset the flip -flop. The 6H pulse output, at the vertical
rate of the system, is fed to the fm reference module, where it is used to
control the white insert function (Chapter 5).
The tape sync processing circuitry for both Ampex and RCA systems is
described further in Chapter 7 under Servo Functions.
6 -6. THE INTERNAL REFERENCE CIRCUITRY

The internal reference module provides gating pulses to the tape sync
processor module, to the tonewheel modulator module, to the fm switcher
module, to the ATC error detector module, and to the burst processor
module. In addition, the internal reference module also provides guide error
to the guide servo module, a lock-sense level to the ATC error detector
module, and a sync fault level to the horizontal AFC module. Refer to
Fig. 6 -8.
The afc sample pulse generator is triggered by gated afc tape horizontal
sync from the tape sync processor module when the tape recorder is locked
(or proper phasing is attained between tape sync and the internal oscil-
lator). At nearly the same time, a trapezoid is generated by the output of
the horizontal -rate multivibrator (oscillator) . Comparison is made at the
first sampling quad, and a dc error voltage is produced.
A second sampling occurs approximately four microseconds after the
first sampling. The second sampling is provided to ensure that transients
do not correct the oscillator erroneously. For this reason, the sync-fault
circuit is permitted to function during the 600 nanoseconds before the
second -sampling circuits start to operate. If it turns out that a sync fault
has been detected, then the second sampling does not occur, and the oscil-
Er :is
a =
-r-
ES
-- ^
o
Y-T-L
8
--1,
fi c;cGD
ÉtiÓ ö5,
Psz
g-- >s E
J®
Ó rarfg
t
i4
Fig. 6-8. Simplified diagram of internal reference, RCA TR -70 system.

lator remains undisturbed. If a sync fault is not detected, then approxi-

mately five microseconds after the afc tape horizontal pulse arrives, the
second sample pulse is generated. This pulse opens the second sample quad
and permits the dc error voltage to be transferred to the oscillator correction
circuits.
The lock-sense circuit determines the phase relationship between the afc
tape horizontal signal and the output signal of the horizontal -rate oscillator.
If the two signals are not in phase, the lock -sense circuit produces a control
signal which disables the sync -lockout circuit and enables the error clamp
circuit in the ATC error detector.
Sync -
Processor -
TP2
AFC Tape Horiz Horiz Rate
Osc Output
Q10 Emitter
4 ps
016 Col lector ISwitcha hie I
r
---4.6 ps
h
AFC Trap
AFC Tape Horiz

P1 -17
Delayed
Tape Horiz
-I II- 4-5 ps
A4-11
IJ
-I I-- 3.2 ps
Q15 Collector
1-2ps
Pl -29 Video -Channel Switch
1
P1-30 8 ps Trailing -Edge Gate
SwitchaOle)
9 ps-I
PI-32 Leading -Edge Gate
I- 3.2 ps
PI-27 AFC Hori./
j-4ps
PI-28
Head Switch Suppress
Fig. 6 -9. Timing diagram for internal reference module.

The output pulses are delayed from the input horizontal-sync pulse by
various amounts determined by their functions. Refer to the timing diagram
in Fig. 6 -9.
Internal reference functions in both Ampex and RCA systems are dis-
cussed further in Chapter 7 under Servo Systems.
6 -7. SIGNAL PROCESSING

The output of the demodulator unit of a tv -tape system can produce
satisfactory pictures on local monitors, but it is not suitable for distribution
to transmitting facilities without further processing. The function of the
signal -processing system is to restore all components of the video signal to
ensure that the output signal meets industry standards and is suitable for
transmission.
The signal -processing system in the RCA TR -70 system (Fig. 6 -10) con-
sists of the following modules:
Video processor module ( 432 )
Video output module (433)
Vertical advance module (532)
Sync logic module (533)
Horizontal afc module (534)
When the composite video signal enters the signal -processing modules,
the sync pulses may be deformed with noise spikes and switching transients
during the blanking interval. The output of the signal -processing modules
is a composite video signal that includes regenerated horizontal and verti-
cal sync pulses, regenerated color burst (phased with the chroma signal) ,
and a clean and adjustable black level. The clean horizontal interval
eliminates misclamping in stabilizing and transmitter amplifiers, which
tends to occur due to the noise in the area below the black level. Mono-
chrome and chroma balance are established so that proper proportions of
each are maintained to produce a good color signal at the output.
As illustrated in the simplified block diagram of Fig. 6 -10, only three
signals (excluding control and standards buses ) are fed to the system.
These are (1) the composite video signal from the CATC video module
(620) , where automatic timing correction has taken place; (2) regenerated
color burst from the burst processor module ( 632) ; and (3) house sync
from the local station sync generator. The output signals from the signal-
processing system include three composite video signals ready for trans-
mission and two video signals used within the tape recorder for monitoring
purposes.
The signal -processing system consists of five modules and is further
separated into two subsystems, the pulse- processing and video- processing
subsystems illustrated in Figs. 6 -11 and 6 -12. These diagrams show, in
simplified form, the general function of the modules.
r NO n
U > > >
ó m Ñ
E 8"-
E,"
oapiA passaooud ya
o^'
o
O N
áá
-6
>
>
ow(S ua6aa au(c ua6aN
a
E
U
3,"
01
2 CO
y
>
E
U
U
HZ
aA >
ó
2 E
U
OS Ind l-e1S
Fig. 6 -10. Block diagram of signal processing system.

Pulse Processing
The three pulse- forming modules that carry out the sync timing and
regeneration functions (Fig. 6 -11) are the horizontal afc module (534) ,
the sync logic module (533), and the vertical advance module (532).
These three modules are closely interlinked functionally and physically to
produce composite blanking, gated horizontal sync, and regenerated (out-
put) sync. It is important to note that the blanking pulses needed for both
monochrome and color tape-recording systems must be regenerated from
the tape sync pulse and hence must be advanced in time relative to the
leading edge of sync.
Start Pulse
Clamped composite video from the video processor module is fed to the
horizontal afc module. Vertical sync is separated from the composite video
and integrated. The integrator circuit has a long time constant and only
functions during the vertical -sync interval. The output of the integrator
produces a start pulse which is coincident with the leading edge of the
second vertical -sync pulse. The start pulse is fed to the vertical advance
module, where it starts the field -rate counter. This circuit counts pulses from
the master oscillator in the horizontal afc module. These pulses occur at a
rate of twice the horizontal line frequency (2 X 15,750 = 31,500) and
therefore with a separation of one-half the tv line time (63.5/2 = 31.5
microseconds) .
Gated Sync Separator

The same clamped composite video that is used in the start -pulse circuit
is used in the gated sync- separator circuit. Gating is performed by the
3.5H pulse from the vertical advance module, the 9H pulse from the sync
logic module, and the horizontal -blanking pulse generated within the
horizontal afc module. This gating of sync performs noise immunity, thus
reducing the chance of transients in the sync -regeneration system. The
output of the gated sync separator is composite sync which is fed to the
sync logic module and used as tape sync in the sync -selector circuits.
Another output from the gated sync separator is fed to a horizontal -rate
multivibrator that removes the double -rate pulses during the vertical inter-
val. These pulses are fed to a sample -pulse generator. Sample pulses of
opposite polarity open and close the switch circuit in the afc loop.
AFC Loop
The purpose of the afc loop is to reference the afc sync pulses, or more
specifically, the output of the master oscillator, to the tape sync pulses.
The master oscillator is a free- running astable multivibrator and is con-
trolled by a dc error voltage which results from sampling the slope of a
trapezoidal voltage waveform. A trapezoidal waveform permits tighter con-
CO
E
U
%.
CL
Fñ
U
CL
á
Fig. 6 -11. Simplified block diagram of pulse processing.

trol, since for a small change in time a large change in voltage is produced.
The trapezoidal waveform is produced by gating the master oscillator and
horizontal -blanking pulses and feeding this pulse to a trapezoid generator.
The sample pulse is produced by the separated composite tape sync.
The sampling medium is a diode quad which acts like a normally open
switch that is closed by the sample pulse. During the pulse time, the closed
switch applies a voltage determined by the height of the trapezoid slope at
that instant to a long- time -constant RC circuit. This RC circuit smooths
out rapid timing variations so that the oscillator is not constantly shocked
by quick changes. During times when the tape sync and afc sync pulses are
not locked, the time constant is much faster, and the master oscillator
responds to changes more rapidly. This permits a faster lock -up of the afc
loop.
AFC Sync Pulses

The afc sync pulses result from gating the delayed output of the master
oscillator with the horizontal- blanking pulses. When the sync -selector
switch on the sync logic module is in the Regen position, the afc sync
pulses are in phase with the tape sync pulses and are used to produce new
or regenerated sync. When the sync - selector switch is in the Proc position,
the afc pulses trail the tape sync pulses. Tape sync pulses in this case are
used for the primary sync source, and the delayed afc sync pulses are used
in case of tape -sync dropout.
If the afc loop is not locked, the afc pulses are not produced, and tape
sync is the sole source of sync. All afc pulses that are coincident with the
video switching point are gated out by the afc - sample suppressor pulse
generated in the fm switcher module. Thus, timing errors in the afc loop
due to mistiming of the leading edge of sync during switching are elimi-
nated.
Blanking
Composite blanking is made up of horizontal blanking from the hori-
zontal afc module and vertical blanking from the sync logic module.
Composite blanking is applied to the video processor module so that old
sync may be removed from the composite video signal and new sync added
later in the video output module. The horizontal-blanking pulse is used
internally in the horizontal afc module to gate the delayed afc sync pulses.
Sync Selection
Three sources of composite sync from the sync logic module are available
to be added to the video output. They are (1) regenerated sync, which is
new sync made from the leading edges of the afc sync pulses from the
horizontal afc module, (2) processed sync, which is made from the leading
edges of tape sync pulses, and (3) house sync, obtained directly from the
local station sync generator.
Sync selection is made by a three -position rotary switch on the front

panel of the sync logic module. In any selected sync mode (Regen, Proc,
or House) , automatic switching and detection takes place so that if the
primary source of sync should fail, a back -up source fills in the gap.
In order to use house sync for the primary sync source, the house sync
and incoming tape sync must be in phase. This condition prevails whenever
the tape-recorder servo system has achieved Pixlock operation and the
phase is set by the horizontal- position control on the play panel.
Automatic sync switching is accomplished by an electronic switching
circuit in the sync logic module. Automatic vertical- advance detection is
accomplished by multivibrator action. If the vertical- advance detector fails
to detect the vertical- advance pulse (3.5H), tape sync is selected as the
composite sync source, and composite blanking, gated horizontal sync, and
the clamp pulse to the video processor module are inhibited.
Clamp Pulse
Stripped tape sync, from the video output module, and horizontal blank-
ing, from the horizontal afc module, are fed to the clamp -pulse generator.
The output of the generator is fed to the video processor module where it
clamps the video circuits.
Video Processing
The two video processing modules that form the video adder and sync
adder functions are the video processor module (432) and the video output
module (433) shown in Fig. 6 -12. These two modules are closely located
physically so that the video signals are not affected by external noise.
Processed Video
Composite video from the CATC video module is applied to the video
processing module, where it is amplified. The gain of the amplifier is
controlled either on the front panel of the video processor module or on a
remote panel.
The amplified composite video is split into a monochrome channel and
a chroma channel. The clamp pulse generated in the sync logic module is
applied to the back porch of the horizontal interval in the monochrome
channel and clamps it to -10 volts. At this point, composite blanking is
added, and the old blanking and sync are removed. A new pedestal level
is added; it also is variable from a remote panel, if selected.
Meanwhile, the chroma portion of the video signal is applied to a clamp
circuit which removes the old burst signal only when the tape recorder is in
the play mode. Chroma gain is adjustable only at the local location on the
front panel of the video output module.
The monochrome and chroma channels, along with the regenerated
color -burst signal, are combined in the video adder. The output of the video
adder (at this point video without composite sync) is applied to the video
THE VIDEO PROCESSING SYSTEM lhl
o- 0 0 0
ó5
4-1
i
O ó
roi
o
E ó
L 5¢
V
.dFF..
C
Ö m
Fig. 6-12. Simplified block diagram of video processing.

output module. In the E -E mode, the new burst is not generated; therefore,
a burst -notch circuit is provided so that the old burst is not eliminated by
the clamp in the chroma channel. The video at this point is called
"processed video," and it is delayed by 0.5 microsecond in order to match
the delays imposed by the sync channel. The chroma portion of the video
signal is fed through a high -frequency boost circuit which compensates for
any loss that results from signal handling.
Sync Adder
Composite sync from the sync logic module is applied to the sync adder
circuits. Sync is added to the processed video at the output of the video
line drivers. Sync gain may be varied either locally on the front panel of
the vertical advance module or from a remote panel. The composite video
signal is supplied to the picture monitor and waveform monitor (cro) .
Composite sync may be removed from the composite video signal in one or
all video output channels (1, 2, or 3) by opening the sync switches, which
are located in the video output module.
International Standards
Most of the circuitry in the three pulse- forming modules consists of
simple time -constant changes in order to obtain the correct pulse widths
needed for the different tv line standards. The verical advance module
contains the circuitry necessary to change the timing of the advance verti-
cal pulse (3.5H) so that it is compatible with 405- and 625 -line standards.
This circuitry includes an additional transistor stage, a miscount amplifier
for driving the first counter when on 625 -line standards.
The switch used in selecting the different tv line standards is located
directly beneath the picture monitor. For operation on 625 -line standards,
the center frequency of the bandpass filter in the chroma channel and the
band -rejection filter is changed from 3.58 megahertz to 4.43 megahertz by
the action of the line standard switch.
In machines equipped with PAL, when 625 -line high -band color opera-
tion is selected, the regenerated burst switches between 0 and 90° every
other line. The regenerated burst moves back and forth with the meander-
ing vertical interval according to the PAL - system specifications.
6 -8. REMOTE VIDEO GAIN
Remote control of various signal amplitudes has been mentioned several

times. A description of the remote video gain control for the Ampex
VR -2000 will serve to illustrate the principles involved.
Etched board 18 (Fig. 6 -13) provides remote control of video gain by
the use of dc voltages instead of by running long video lines to the remote
control point. Thus, the effects of external disturbances on long video lines
in the control circuit are eliminated.
5
v a
>
o ._
.-vv\--1 H,,
"lQ.Q.p 1
G°
E C<
>
_ +N//- Ñ
ce
>
N
0
z
Fig. 6 -13. Remote -video -gain etched board 18.

The incoming video signal from J008s enters the circuit board at pin R
and is coupled by emitter follower Q6 through resistor R15 to the junction
of diodes X1 and X4. The anodes of these diodes are connected to the
emitters of Q2 and Q5, respectively.
The input impedances of diodes X1 and X4 and transistors Q2 and Q5
are the key to the remote control of video gain. The dynamic impedance
of any semiconductor device is a function of the dc current through it. The
resistive component of the diode impedance is approximately 25 divided
by the dc current in milliamperes. Thus if the dc current is 1 mA, the
resistive component is 25 ohms; if the current is 5 mA, the resistive com-
ponent is 5 ohms.
The impedance at the junction of R15 with the cathodes of X1 and X4
is very low (i.e., of the order of 10 to 20 ohms) . The resistance of R15
(680 ohms) is considerably larger than the junction impedance. It thus
establishes a constant -current source that is relatively independent of any
current changes that occur in later circuitry. Fig. 6 -14 is a simplified
diagram that illustrates the mechanism involved. At the junction of R15
with diodes X1 and X4 there is a variable resistance to the left and a
variable resistance to the right. The constant current from Q6 divides
between the two grounded -base amplifiers (Q2 and Q5) in the inverse ratio
of their dynamic impedances.
The output of the circuit of Q2 and Q5 is taken from the collector of Q2.
When Q2 saturates, Q5 cuts off, and all the dc current through R15 passes
through Q2 and R12. This is the condition for maximum gain. When Q2
cuts off, all the current through R15 passes through Q5 and R8. This is the
no- signal condition.
If transistors Q2 and Q5 were used by themselves and their conductions
varied differentially, the dc voltage at the collector of Q2 ( the output)
would follow the changing dc level, causing an unwanted "bounce" in the
resulting gain. Such a sudden change of dc level would seriously disturb
subsequent clamp circuits in the processing amplifier. Transistors Q3 and
Q4 provide a dc current change which is (ideally) equal and opposite
+12 V
R8
Fig. 6 -14. Simplified diagram

I1F ~I2 of remote control of video gain.
Series Impedance Series Impedance
1X4 -Q5 Emitter) (X1 -Q2 Emitter)
R15 6800
Video Voltage Generator
Q6
-12 V
to the "bounce" current that appears in R12. The dc current through R12
remains nearly constant throughout the range of video gain. The bases of
Q3 and Q5 are connected through R14; the collectors of Q4 and Q5 are
common. Thus Q3 and Q4 act to cancel the dc current and tend to hold the
collector of Q2 at a constant dc voltage. (Neither Q3 nor Q4 is concerned
with the video signal.) The network consisting of R4 and C3 (Fig. 6 -13)
provides dc feedback to further reduce the unwanted gain "bounce."
When relay K1 is de- energized, panel control R003B (VIDEO GAIN)
controls the dc bias of Q1 and Q3; when K1 is energized, an identical
potentiometer on the remote control panel replaces R003B.
The range (or calibration) of R003B is set by R18. Control R20 adjusts
the base bias of Q4 to balance the currents through Q3 and Q4, and to
equalize their currents with the currents through Q2 and Q5. An RC net-
work that includes R14 reduces differential phase and differential gain
and maintains the frequency response through etched board 18. Inductor
L1 peaks the frequency response of the collector output of Q2. Inductor L2
isolates feedthrough of the video signal appearing at the collectors of Q2
and Q3. This is done to prevent high- frequency loading of the signal by
the emitter of Q4.
The output of Q2 is coupled by emitter follower Q1 through response -
equilizing control R1 to pin M, from which it is routed to the remote
position of switch S2002Á ( VIDEO OUT) The network composed of R1,
.
R2, and Cl compensates for the capacitance presented by the unterminated

cable that is connected to pin M.
The connections between the remote video gain panel and etched board
18 are made through a 10- conductor cable to J015 p (REMOTE) This cable
.
carries only dc control currents and voltages, and it may have any reason-
able length.
EXERCISES
Q6 -1. What is the basic signal error in the playback of a tape when the guide
positioning is not identical to that used for recording?
Q6 -2. What is the basic element in any quadruplex -tape ATC system?
Q6 -3. How is the element of Q6 -2 basically used for time -error correction?
Q6 -4. With the tape system in the nonsynchronous mode of playback, what
are the basic signals used for comparison to generate an error signal?
Q6 -5. With the tape system in the synchronous playback mode, what are the
basic signals used for comparison to generate an error signal?
Q6 -6. What is the normal correction range (in nanoseconds) for automatic
time correction circuitry?
Q6 -7. Are the varicaps in an electronically variable delay line forward or re-
verse biased?
Q6 -8. What signals are used for comparison to generate an error signal in
color ATC systems?
CHAPTER
The Servo System
The servo constitutes the synchronizing section of the television -tape

system, as contrasted to the video section. The television -tape servo system
is electromechanical in nature; the input is an electronic signal, and the
output device is an electric motor with a mechanical load. Three basic
servo systems are involved: The rotating- head-motor servo, the capstan -
motor servo, and the vacuum -guide -motor servo.
NOTE: The entire combined headwheel (drum) and capstan servo that
makes it possible to lock one or more recorders to any external sync source
or to each other is termed the Intersync section of Ampex recorders. This
is equivalent to RCA's Linelock or Pixlock sections. See Table 6 -1 (Chap-
ter 6) for the correlation of terminology.
7 -1. RELATIONSHIPS
The basic synchronizing function of the tv -tape system will be clear if
the following basic relationships are clearly visualized:
1. Relationship of the head timing pulse (tachometer) to the video head
recording vertical sync.
2. Phasing of the control -track signal being recorded with the video
tracks being recorded, as timed from the 60 -Hz reference pulse from
which the 30 -Hz frame (edit) pulse is derived.
3. Relationship of the tracking control to the video head playing back
vertical sync.
Fig. 7 -1 illustrates the factor that determines which of the four video
heads records the vertical sync. Since the 240 -Hz signal derived from the
rotating-head shaft is compared in phase with the 240 -Hz signal derived
from vertical sync (reference pulse) , the particular head in contact with
the tape at the leading edge of this pulse is the one that will be recording
vertical sync. Thus, the physical orientation of the timing ring of the tone -
186
Timing Pulse
From Head Speed
J L Timer
ITachometerI
Output Feedback
AA -This Coincidence Related

to Particular Video Head
r----1
I
Motor
Tape
in Contact With Tape
Closed -Loop
rt
Servo
I
240 Hz
I
60 -Hz Ref
Pulse Derived
From Vert Sync
160 Hz
x4 -
?40 Hz
Ph se
Comp rison
240-Hz
Osc
Fig. 7 -1. Relationship of tachometer signal to head in contact with tape.
wheel notch relative to the head assembly determines which of the four
heads will record vertical sync.
The notch in the RCA tonewheel is directly over the tonewheel head
when video head No. 1 is in the approximate center of the tape. Thus
video head No. 1 records vertical sync in the RCA machine. The orientation
of the Ampex tachometer is such that the leading -edge reference occurs
when head No. 4 is in the approximate center of the tape. Thus, head
No. 4 records vertical sync in the Ampex machine.
The A -scope ( which monitors the playback channels) on the Ampex
control panel is a driven type, with the horizontal sweep being triggered
from a derivative of the timing tachometer. The output of the four video
heads during one revolution of the head drum is displayed on one hori-
zontal sweep of the oscilloscope. With the tracking control centered (same
head playing back vertical sync as recorded it) , the presentation is as shown
in Fig. 7 -2. Since vertical sync occurs at the approximate center of the
channel-4 track, the retrace is triggered at this time, and channel 4 con-
tinues a short distance on the left section of the trace.
NOTE: It should be understood that any existing difference with respect
to which video head records vertical sync does not affect compatibility or
interchangeability of tapes between any two systems. When a recorded
II2
Fig. 7 -2. A -scope presentation of
Ampex head switching.
4
III
2
IIII E i pi IIII IIII
Triggered at Time of Vert Sync Here

iApprox Center of Channel 41
tape is played back on a different head than that used for recording, the
tracking control is adjusted to obtain the best match of head-to -tape
response.
The phasing of the control -track signal, recording of the video tracks,
and frame (edit) pulse is illustrated in Fig. 7 -3. (Since the control -track
signal is used to provide a timing reference during playback for the
capstan servo, it is sometimes called the tracking- control recording.)
The 240 -Hz sine wave on the control track represents the current in the
control -track head during recording, and hence the degree (and polarity)
of magnetization on the tape. The current nodes (zero magnetization)
occur in phase with alternate video -track guard bands. Thus, current nodes
and current maximums occur in phase with alternate video -track guard
bands. The control-track sine -wave current is so phased that the edit pulse
(30 -Hz frame pulse derived from the 60 -Hz reference pulse and occurring
every 33 video tracks) occurs at the point of maximum current in the
guard band between the second and third tracks following the track that
contains vertical sync. This is between tracks 3 and 4 if track 1 contains
the vertical sync, as shown in Fig. 7 -3.
It is important to note that proper phasing of the control -track signal
with the edit pulse (and hence the reference pulse) automatically estab-
lishes a standard phasing relative to the video tracks being recorded; this
is most important for interspliceability. If a tape is spliced onto another
segment which does not have identically the same phasing between video
tracks and control tracks, the tracking control must be manually readjusted
immediately after passage of the splice.
Tape Travel 00
90° 180° 270° 360°
Tape
Head
Edit
VScans
Pu se Tachometer
T
Gen
Video 30 Hz 240 Hz
Tracks
Start of Vert
During
Recording - 240 Hz
Rotating Heads
/
Blanking 30 Hz
On ly
End of Vert
Control Trac\
Head
Blanking Playback
Current Nodes
(Zero Magnetization)
Capstan
.4 -411. Motor
Capstan
Fig. 7 -3. Phasing of video tracks, control -track signal, and edit pulse.
Fig. 7 -4 shows the relationship of the tracking control to the video head
playing back vertical sync. As shown, either the control-track signal or the
head -tachometer signal may be shifted in phase so that any one of the four
video heads will track on vertical sync. Ampex machines track by means of
the control -track signal, and older RCA machines track by means of the
tone -wheel signal. All older RCA systems modified for synchronous opera-
tion and all later-model RCA recorders also use the control -track signal
path for tracking control.
Tape
Variable
240 Hz
' Phase
Older RCA
"Tracking Control" May Be Here or Here
Rotating Heads
240Hz Variable Phase Capstan To Capstan

Control-Track
Head : Ampex and RCA Servo Power Amp
Fig. 7 -4. Tracking control of video head playing vertical sync.
7 -2. BASICS OF THE AMPEX INTERSYNC

SERVO SYSTEM
The magnetic -tape recording and reproduction of television signals in-

volves the scanning of a moving tape by four video heads that are mounted
in precise quadrature on the periphery of a rapidly rotating head drum.
The longitudinal movement of the tape is synchronized with the rotation
of the head drum by the action of two mutually synchronized electronic
servos. In the basic system, tape speed is controlled by the action of the
capstan servo system; head -drum rotation is governed by the head -drum
servo.
A better understanding of the Intersync servo functions and circuitry
will be gained by a brief review of the standards of the television industry
and how they relate to problems accompanying the achievement of a high
degree of servo synchronization and timing stability. The discussion which
follows deals with the:
1. Elements of the television signal

2. Limitation of timing errors
3. Servo record -mode functions
4. Servo play -mode functions
5. Servo control modes
lß)(ì TELEVISION BROADCASTING
Elements of the Television Signal

To produce a complete frame of a television picture, the raster is com-
pletely scanned twice by the electron beams in the camera and the picture
tube. Each scan starts at the top of the picture and ends at the bottom and
is composed of a series of closely spaced horizontal lines; the horizontal
paths traced by the electron beam on the second scan are interlaced mid-
way between those traced on the first scan.
The television standard in use in the United States is 525 lines per frame
and 30 frames per second, which requires 60 scans per second. When each
scan is completed, the scanning beam is blanked out and returned to the
top of the raster by a synchronizing pulse occurring at the vertical rate,
which is 60 times per second. These vertical- synchronizing pulses are
recorded on the tape midway across the area occupied by the video informa-
tion, and they recur on every sixteenth path of recorded video information.
The vertical pulses that initiate the first scan of a frame coincide in time
with a 30 -pps frame pulse that is recorded simultaneously on the longi-
tudinal control track at the lower edge of the tape. The frame pulse re-
corded on the control track always appears on the tape in the same relative
position with respect to the corresponding vertical -sync information re-
corded as a part of the composite video signal.
The 525/30 television standard uses 525 horizontal traverses of the
raster by the electron beam for each complete frame. There are therefore
15,750 such traverses per second, requiring 15,750 horizontal pulses per
second. This is known as the horizontal rate. When video information is
present during the recording, the space following each horizontal pulse,
with one exception, contains video information corresponding to a full
horizontal line in the picture.
The vertical- synchronizing information that initiates a new frame is
always preceded by a half-line of video information; the vertical- synchro-
nizing information that appears on the sixteenth path following completes
the frame and is always preceded by a full line of video information.
Thus, complete frames are presented 30 times per second. Fig. 7 -5 shows the
vertical- synchronizing information which appears at the middle of every
sixteenth transverse path. This information may be seen to include (in
sequence) six equalizing pulses, six vertical -sync pulses, and six equalizing
pulses. The recorded 240 -Hz control -track signal and the superimposed
30 -pps control -track frame pulses appear at the bottom edge of the tape.
Fig. 7 -5 also shows that the stationary control -track record/reproduce head,
which is mounted on the video head assembly, is located 3/4 inch "down-
stream" from the plane of the video heads. The control-track head records
and reproduces the 30 -pps frame pulses in addition to the 240-Hz control -
track signal. Because of the "downstream" location of this head, the vertical -
sync information associated with each frame pulse is 45 video tracks
"upstream" from it.
r-
..
. . .. . . __.--Imo . . r
/ Tv:
O
O
O
E
-ó
Ñ
v
m
rá
ó d
vc E -8
c
°
Á
LL
«t. iw
Ó
a ¡
...... __..._
Ó
rTo
/
I . . .. _____
.
rl jv
o
,;§
. . . .
.....
ÿ N
-
. ... ,- -___-......
.. ... ....._ ....
Fig. 7 -5. Positional relationship of video and sync on coated surface of tape.
1 92 TELEVISION BROADCASTING
Limitation of Timing Errors

There is a minimum degree of inherent timing error in the record/
reproduce process. Because of this, any recording will contain a normal
amount of timing errors to which the playback process will add more.
Timing errors originate from a variety of causes, of which the following
are the more frequently encountered:
1.Momentary changes of head -drum velocity caused by variations of
head-drum loading.
2. Momentary phase transients in either the reference signal or the
video signal.
3. Momentary loss of horizontal or vertical sync.
4. Instability of the reference signal caused by a malfunction of the afc
circuitry in the sync generator.
The principal function of the Intersync servo unit is the maintenance of
synchronization between the reproduced video signal and other video
Reference Vertical and Horizontal -
LiPhase
Sync Horizontal Phase
Separators Modulator
Pulses Comparator
Proc -Amp Vertical and Vertical Drum-

Sync Horizontal Phase Drum
Motor
Pulses Separators Comparator Drivers Motor
r
Drum -
Tach
Drum -lach
Sh per
Dr um-
Phase
I -0--
Drum
Oscillator
Input Comparator o
Electronic Switch
Control -Track Capstan

Record Capstan
Motor
Am pl if ier Drivers Motor
Control -
Record
Control-Track
i
Track Playback Phase Cap tan
Comparator Oscillator
Head Play Amplifier
Fig. 7 -6. Simplified diagram of Ampex Intersync servo system.
sources to a degree that permits program switching between these sources

without interruption of picture continuity. The Intersync servo unit accom-
plishes this by locking the reproduced signal to the sync reference to which
the other video signals are locked, and by holding the timing errors in the
reproduced signal within very narrow limits. Fig. 7 -6 shows a simplified
block diagram of the Intersync servo system.
Servo Record -Mode Functions

During the record mode, the Intersync servo unit provides two closed
servo loops that control head -drum rotation and maintain maximum timing
stability of the recorded signal. The positional loop (Fig. 7 -7) determines
the average rotational rate of the head drum and the instantaneous angular
position (or phase) of the video heads with respect to the reference signal.
Its gain and bandwidth are designed to provide a "soft" servo control that
resists the reaction of the head drum to high -rate -of- change disturbances
that may appear in the video or reference signal. Control is maintained by
the phase (i.e., time) comparison of the tachometer signal (representative
of head -drum phase) with the sync components of the incoming video
signal. The resulting measure of timing error is represented by a propor-
tional voltage that controls the frequency of the drum -oscillator output.
This action places video head 4 in position to record the vertical -sync
-
information at the precise center of the tape width.
Incoming
Video Sync-1'o Timing Voltage
Phase
o Error Controlled
Modulator
Mode °o Detector Drum Osc
Selector
Switch
1
Tach Amp 1
Drumer
ad Dr
Sh per Head -
Dr um
Motor
Tach Signal
Fig. 7 -7. Block diagram of positional loop.
If a program transition to a nonsynchronous video signal source is re-

quired during the making of a recording, the positional servo must correct
the angular position of the head drum at a controlled rate, in order to
minimize a corresponding instability that will occur during the subsequent
reproduction. In the event of severe interruptions of the video sync com-
ponent, the Intersync servo unit may be switched to the power -line or
external reference position.
The second loop (Fig. 7 -8) provides damping and is used during all
modes of operation. Its damping action minimizes high- rate -of-change dis-
turbances that affect the instantaneous frequency of head-drum rotation.
Such disturbances include those resulting from the natural tendency of the
drum motor to "hunt" at a rate of 7 to 10 Hz, or from momentary changes
of head -to -tape pressure (drum loading) caused by the passing of a tape
splice. The minimizing of these disturbances establishes a flat frequency
characteristic within the bandpass of the servo system.
In the damping loop, a frequency- sensitive detector derives a signal

voltage that is proportional to any instantaneous changes in the frequency
of the tachometer signal. This signal voltage is applied to a phase modulator
which also receives the drum -oscillator signal. The oscillator signal is modu-
lated in terms of corresponding momentary phase shifts to counteract the
disturbance, and then it is routed to the drum -motor driver.
Output
Phase Drum -Motor
From Drum
Modulator Driver
Oscillator Head -
Drum
Motor
Tachometer
Frequency
Shaper and Tach Signal
Discriminator
Amplifier
Fig. 7 -8. Block diagram of damping loop.
During the record mode, the capstan servo system maintains constant
velocity of the tape in its movement past the rotating video head drum.
The frequency of the capstan drive signal is 60 Hz (U.S. system) derived
from an oscillator and locked to the drum-tachometer signal. The capstan
synchronous motor is thus caused to drive the tape at a nominal rate of 15
inches per second (U.S. system) but is electronically locked to the rotation
of the video head drum.
Derivation of the capstan drive signal in this manner, rather than directly
from the power line, prevents power -line phase transients from having an
effect because of the "soft- lock" characteristic of the drum servo loop. In
addition, if the head drum changes speed momentarily to rephase, the cap-
stan changes speed proportionally and thus maintains a constant interval
between the recorded transverse tracks across the width of the tape.
Servo Play -Mode Functions
During the making of a recording, the tape moves longitudinally at the
nominal rate of 15 inches per second, while the video heads rotate at the
rate of 14,400 rpm (US system) The carefully controlled positional
.
relationship between vertical -sync information and the 30 -pps frame pulse
is established at this time, during which the tachometer scans the rotation
of the video head drum and generates one complete cycle of a square wave
during each revolution.
Because the drum turns at 240 rps, the square wave produced by the
tachometer has a frequency of 240 Hz and is in exact phase with the drum
position. The tachometer signal is recorded on the control track at normal
level together with the 30 -pps frame pulse, which is recorded at the level
of tape saturation. The considerable difference between the recorded levels
of the frame pulse and the control track permits positive identification of
both reproduced signals.
Reproduction of the recorded composite video signal requires accurate

servo controls that will position the tape longitudinally and phase the
video heads transversely to re- establish the positional relationship existing
during the recording process. The tape is positioned longitudinally by the
capstan, which is controlled by comparison of the reproduced control -track
signal with a reference signal. Simultaneously, the phase of the video head
drum is precisely indicated by the phase of the tachometer signal, which is
compared with that of the reference signal. An error in the longitudinal
position of the tape or in the phase of the head drum results in a correction
voltage that acts on the appropriate servo system to cancel the error.
Servo Control Modes
The Intersync servo unit allows manual selection of five operating
modes, as follows:
1. Automatic: In record, this mode locks the recorder to the incoming
video vertical -sync signals. In play, it permits full Intersync operation
which tightly locks the recorder to reference horizontal and vertical
sync.
2. Vertical: This mode locks the recorder to the video vertical -sync
signal in record and to reference vertical sync in play; it maintains
accurate frame lock when no -roll switching is desired.
3. Horizontal: This mode locks the recorder to the video vertical -sync
signal in record and tightly locks to external reference horizontal
sync in play. Use of the horizontal mode minimizes any effects from
signal discontinuity and permits high quality when playing back
spliced color and monochrome tapes. Properly made splices are not
visible on the screen. The reproduced video will not necessarily be
vertically framed with respect to reference sync.
4. Normal: For monochrome use, this mode locks the recorder to the
signal vertical sync in record and to the power line in play.
5. Preset: In this mode, the Intersync servo unit may be set to any
desired record or reproduce reference. This mode is similar to the
normal mode, except that the timing reference source is selected by
the chassis switches labeled PRESET REFERENCE SELECTORS. The preset
mode is used for checking various circuits and to meet unusual re-
quirements of the user.
Automatic Mode -The Intersync servo unit performs its major function
during the recovery of signal information from a tape recording. To do this,
in the automatic mode it provides servo controls that identify the particular
transverse track in which a frame begins, positions the tape to cause the
video heads to scan the exact center of each transverse track, and phases
the head-drum rotation to cause one particular video head (number 4) to
scan the identified track when switched to track 1. Other heads may be
selected to scan the identified track. Fig. 7 -9A shows a block diagram of the
Off -Tape ontrol -Track

Capstan
Control Track Reproduce
Motor
Signal Head
30 -PPS Frame Pulses) Error Voltage Drive Signal

j 1 t
Voltage Cap tan -
Phase
30 PPS P Controlled Motor
Comparator
Capstan Osc Drivers
Reference Position -Control Signal
Voltage -
Phase Phase
60 PPS Cont oiled
Comparator Modulator
Drum Osc
Drive Signal
Damping Error Signal
Frequency Frequency Drum -Motor

Discriminator Discriminator Drivers
Head
Motor
Tachometer
Amplifier Tach Signati
and Shaper
(A) Initial condition.
Reproduced Voltage Capstan

Phase Capstan
240 Hz Controlled Motor
Comparator Motor
Control Signal Capstan Osc Drivers
Reproduced Vol age

Phase
Horizontal Cont oiled
Comparator
Sync Drum Osc
Reference
Ph se Drum -Motor
Horizontal-. Modulator Drivers Head -
Sync
Drum
Motor
Tachometer
Frequency
Shaper and Tach Signal
Discriminator
Amplifier
(B) Final condition.

Fig. 7 -9. Block diagrams of automatic mode.
automatic mode, initial condition, and Fig. 7 -9B shows a block diagram of
the automatic mode, final condition.
Track identification is derived from the 30 -pps frame pulse recorded on
the control track, which controls capstan rotation to position the tape
longitudinally. During this period, head -drum rotation is locked to a
reference signal (e.g., station reference sync) , but the video heads are not
yet in contact with the tape. Thus the longitudinal positioning of the tape
is accomplished by relatively large changes of tape speed during a period
when recorded vertical- and horizontal -sync information is not being repro-
duced. Because the control -track record /reproduce head is in contact with
the tape at all times, frame pulses are reproduced as soon as tape movement
begins. It should be remembered that each frame pulse coincides in time
with the vertical -sync information that initiates a new frame (Fig. 7 -10) ,
and that the frame pulse is recorded on the control track, whereas the verti-
cal -sync information is recorded on a transverse video track.
H/2 Interval of 1/2 Line

Start of
Vertical Sync Interval New Frame
Equalizing Pulses -H- Vertical Sync Equalizing Pulses Horizontal

Pulses
Tip of
Sync
-If11-_ 1111111
Blanking' HH /21- -H-
Y Third Vertical Serration Vertical Sync
H/2 - Interval of One -Half Line

H Interval of One Line
Peak White Level
Fig. 7 -10. Vertical- interval waveform at end of frame.
After 1 or 3.8 seconds, depending on the type of start -up used, and after
capstan rotation has stabilized or nearly stabilized the longitudinal position
of the tape, the vacuum tape guide brings the tape into contact with the
video heads; the heads then begin reproduction of the recorded video
information. If for any reason the control -track frame pulses are not present
and framing is not accomplished as described above, duplicate framing
information derived from the reproduce video tracks controls the capstan in
the same manner as the control-track frame pulses. When framing is com-
pleted and video is being satisfactorily reproduced, the capstan servo and
head -drum servo change to the same condition as the horizontal -mode final
condition.
Horizontal Mode -In
this mode, capstan rotation is controlled by phase
comparison of the reproduced control-track signal with the tachometer sig-
nal. There is no other correction of the longitudinal position of the tape.
Fig. 7 -11 describes the horizontal mode in the initial condition; Fig. 7 -9B
describes the horizontal mode in the final condition (same as the automatic
mode in the final condition) .
During the few seconds after the reproduced control -track signal appears
and before the vacuum guide engages the tape with the video heads, head -
drum rotation is stabilized by phase comparison of the tachometer signal
with reference vertical sync. At the conclusion of this interval, if repro-
duced sync pulses are being received, the control of head -drum phase is
transferred to the output of the horizontal comparator. It remains under
this control unless the tachometer phase comparison shows a slipping
condition of the tachometer phase with respect to reference vertical sync
or unless the reproduced sync completely disappears.
Reproduced
a, Frequency
Control Track 1
240 Hz Discriminator
Signal
Voltage Capstan -
Phase Capstan
Comparator Controlled Motor
Motor
Capstan Osc Drivers
.---240 Hz
Phase
Comparator
Voltage
Controlled
Drum Osc
Reference
Frequency
VerticalDiscnmmator
60 Hz
Phase
Modulator
Dr um -Motor
Drivers
Sync Head -
Drum
Motor
Tachometer
Frequency
Amplifier Tach Signal D
Discriminator and Shaper
Fig. 7 -11. Horizontal mode, initial condition.
Whenever resynchronization is required by the appearance of momentary

phase transients, or by the momentary loss of reproduced horizontal sync,
it is accomplished by again locking the reproduced horizontal sync to the
nearest reference horizontal -sync pulse.
The horizontal mode is particularly designed for use with Colortec,
which is the Ampex direct color process. Because this mode provides rapid
recovery of lock-in, the interval during which timing errors in the repro-
duced signal exceed the Colortec correction range is minimized. However,
the horizontal mode does not provide frame or field synchronization either
initially or in the restoration of lock-in following momentary loss of sync.
Consequently, program switching between reproduced video and other

video sources will produce brief disturbances in the picture presentation.
Vertical Mode -Frame synchronism within ± 10 microseconds of the
reference is achieved rapidly in two sequential operations. Initially, the
capstan positions the tape to place the beginning of a recorded frame in
line with the head drum coincidentally with the beginning of a frame of
information in the reference signal. This is accomplished by phase compari-
son of a reproduced frame pulse with a reference frame pulse and is
normally completed within 3 to 4 seconds following the appearance of the
reproduced control -track information. During this period, head -drum phase
is corrected, by phase comparison of the tachometer signal, to place video
head 4 at the center of the tape width coincidentally with the occurrence
of reference vertical sync. The block diagram of the vertical mode in the
initial condition is the same as that shown in Fig. 7 -9A.
Reproduced Voltage Capstan-

Capstan
Control -Track Controlled Motor
Motor
Signal 240/250 Hz Capstan Osc Drivers
Voltage
Reproduced Phase
Controlled
Vertical Sync 50/60 pps Comparator
Drum Osc
Reference Ph se Drum -Motor

Vertical Sync 50/60 PPS
Modulator Drivers Head -
240/250 Hz Drum
L__. Motor
Frequency
Discriminator
Fig. 7 -12. Vertical mode, final condition.
At the conclusion of this period, capstan control is transferred to a

voltage derived by phase comparison of the reproduced control -track signal
with the tachometer signal. Tape movement is then locked to head -drum
rotation. Control of head-drum rotation is transferred to a voltage derived
by phase comparison of the reproduced vertical sync with reference vertical
sync. The occurrence of a phase transient or an imperfect tape splice
may cause loss of lock. If this occurs, the capstan will again accomplish
framing as described above, provided the REFRAMING switch on module 14
is set to the Cap position. If the REFRAMING switch is set to the drum
position, the drum advances or retards as many rotations as necessary to
restore frame lock without loss of tracking. Fig. 7 -12 describes the vertical
mode in the final condition.
The TRACKING control on the main control panel has the "home" track
setting marked on its dial. At this setting, video head 4 will reproduce the
transverse video tracks that contain the vertical-sync information.
While a given video head assembly is in service, the TRACKING control
should be kept at the "home" track setting during the recording and repro-
duction of a given tape. This will assure the best possible reproduction of
that tape because the reproduce conditions recreate those present during
the making of the recording.
By altering the setting of the TRACKING control, any one of the other
video heads may be selected to reproduce the tracks that contain the vertical -
sync information. During the reproduction of a recording made by a
different video head assembly (or a different system) , there may be an
occasional subjective improvement of reproduced picture quality when
head 1, 2, or 3 is selected to reproduce these particular tracks.
Normal Mode-The normal mode is recommended for monochrome
reproduction when the recording is known to contain sync- timing discon-
tinuities such as those that occur when a tape is improperly spliced. Re-
produced video is "soft- locked" to the power -line frequency. However,
reproduced sync timing is tc ..ally ambiguous with respect to the station
sync -generator reference. Tape reproduction results without regard to
framing and is based on phase comparison of the 240 -Hz control -track
signal with the drum -tachometer signal, which in turn is phase locked to
the power line. Fig. 7 -13 describes the Intersync servo when operated in
the normal mode.
Reproduced
Control -
Frequency
Track 1 r
Discriminator (Connected Only Before
Signal 1
( Phase Lock Is Achieved)
Capstan
Phase Capstan Capstan
Motor
Comparator Oscillator Motor
Driver
I
Ph se Drum Phase Drum Motor
Tach Signal
Comparator Oscillator Modulator Drivers Head -
Drum
(Connected Only Motor
Before Phase Lock
Power f Is Achieved)
Frequency
Line Tach Signal
60 Hz
Discriminator Damping Error Signal
Reference
Tachometer
Frequency
Amplifier
Discriminator
and Shaper
Fig. 7 -13. Block diagram of normal mode.

Preset Mode -In the preset mode, the controls of the Intersync servo
unit may be set to any desired record and /or reproduce reference. This
mode is used mainly for checking various areas of the circuitry and to
meet unusual requirements of the user. The preset mode is similar to the
normal mode except that its timing reference source is chosen by means
of two chassis- mounted selector switches labeled PRESET REFERENCE
SELECTORS.
7 -3. THE RCA HEADWHEEL SERVO

The headwheel servo system (Fig. 7 -14) provides servo -controlled
three -phase power to the headwheel motor in both the record and play
modes of the RCA TR -70 tape machine. In the record mode, the headwheel
servo tightly locks the motor velocity and phase to vertical sync. This rela-
tion ensures that vertical sync will be recorded by video head 1 at a pre-
cisely determined spot near the center of the tape width and that exactly
the required number of video tracks will be recorded per tv field. On do-
mestic (525 -line) standards, the headwheel rotates at four times the 60 -Hz
field rate (240 revolutions per second) , and, therefore, since each revolu-
tion produces four tracks, sixteen tracks are recorded in each field period.
On international (405- and 625 -line) standards, however, the headwheel
rotates at five times the 50 -Hz field rate (250 revolutions per second) , and
20 tracks are recorded per field.
If, during recording, the headwheel motor velocity were fixed, without
reference to vertical sync, the recorded vertical -sync pulses would drift
across the video tracks, and the recorded edit pulses (see Section 7 -4)
would drift on the control track. Although a tape recorded under these
conditions could be played back successfully on the same machine if no
splices were made after recording, it would not be interchangeable with
tapes made on a different machine.
During playback, the primary function of the headwheel and capstan
servo systems is to control the respective motors so that a stable picture,
with accurate sync and video frequencies, will be produced. Different
requirements, however, such as roll -free switching, special effects, and
monochrome or color ATC, demand a choice in the tightness of control and
in the phasing between tape and external (local) sync. Therefore, during
playback any of four different modes of servo operation may be selected:
tonewheel, Switchlock, Linelock, or Pixlock. These modes are discussed in
the system functional description below.
In both the record and play modes, the headwheel servo functions basi-
cally as a closed feedback loop in which error signals regulate the power
supplied to the motor. These error signals are derived by periodically
measuring the angular velocity and phase of the magnetic tonewheel (on
the headwheel motor shaft) against standard references. During playback
in the Pixlock or Linelock mode of servo operation, however, the system
is expanded into a compound loop containing additional feedback from
d
3 óÓ
CO
CU ai
ó
Uco
5 ó
0
s_
3 - 3
o - ó0 c ú
E n.o E d E =
Y N U
ó á o á
U
U Ó
> O U -==
1001 u0 .11 3 OA10 S MI

dd.73
ó
IOVI u0aJ3 4Dolauil 3H1 CL
ú00
_ Ff
CC
,..
_
>
ú
c
ó
m n
N
c O
CC (.D
NT
Fig. 7 -14. Simplified block diagram of headwheel servo system.

tape sync and the error buses of the monochrome or color ATC system
(Chapter 6) .
The headwheel servo system consists of the following components:

1. Timing and signal processing
A. Reference generator (613)
B. Tonewheel processor (612 )
2. Modulation and motor drive
A. Headwheel modulator (609)
B. Power amplifiers (606, 607, 608)
3. Error detection
A. Tonewheel servo (611)
B. Linelock (610)
The following sections describe the overall operation of the headwheel

servo system during recording and playback.
Record Mode
In the record mode, an afc circuit in the tonewheel processor module,
locked to the pulses from the magnetic tonewheel attached to the headwheel
motor shaft, produces a sine wave at twice the tonewheel frequency. This
sine wave serves as the basic motor drive signal (see Fig. 7 -15) Networks
.
in the headwheel modulator module split this signal into three sine waves
which are 120° apart in phase. Chopper circuits controlled by a dc error
signal from the tonewheel servo module then pulse -width modulate the
three sine waves at a line- frequency (horizontal) rate. The amplitude and
polarity of the dc error signal control the modulating pulse width, and,
since the drive -signal amplitude is zero during the pulse interval, the pulse
width determines the energy remaining in each cycle of the sinusoidal
signal after modulation. For example, a negative error signal of maximum
amplitude reduces the pulse width to zero and leaves the sine wave in the
unmodulated, or maximum- energy, condition. Conversely, a positive error
signal of maximum amplitude increases the pulse width to a full horizontal
line period, thereby leaving zero energy. Filters in the modulator output
remove the line -rate information and leave three sine waves, separated in
phase by 120 °, with amplitudes depending on the energy in the pulse -
modulated signals. Separate power amplifiers (modules 606, 607, and 608)
are driven by each of the three headwheel- modulator output signals and
provide just enough power to drive the three -phase synchronous motor at
one -half the synchronous speed.
As the headwheel rotates, the tonewheel pickup head provides one pulse
per headwheel revolution. These pulses provide feedback to a velocity -error
detector and a phase -error detector in the tonewheel servo module. The
velocity -error detector compares the period of the tonewheel pulses with
an independent fixed reference, which is the combined period of two free -
running multivibrators. The total multivibrator period is one-quarter of
-i
3
n
ao E E
ñ
, E
x,f
N
Q
J 2ó
O O
u
Ó
ú
r o
O
L n
pV a_
Ó
C
U
CL
y u ó
> ÿ CO J
CC N 0 c 2á c
$
8 i ó cá S
p
Op m
C U
V1 J U
_11¿ N CL
E
c'D
ÑCe N
Ho
u
c
O
W
L
oC
L
ry
_ - J
u
VI
Ñ
--a-.1
Ñ
d 3
cd
_
p
1.
N
Ñ
vÑ
>
Só
m ó
n
ó
Fig. 7 -15. Headwheel servo system during recording mode and tonewheel or
Switchlock playback mode.
the tv field period (1/240 second) on domestic standards, or one -fifth of

the field period (1/250 second) on international standards. Therefore the
velocity-error signal depends on the difference between the tonewheel
velocity and 240 (or 250) revolutions per second. The phase -error detector
compares the phase of the tonewheel pulses with the phase of a reference
vertical ( field -rate ) signal provided by the reference generator module.
In the record mode, a switch on the reference generator module permits the
selection of either external ( house) sync or sync separated from the in-
coming video signal as the source from which the reference vertical signal
is derived.
The velocity- and phase -error signals are combined in a manner which
causes the velocity signal to predominate when large errors occur and the
phase signal to predominate when small errors occur. Consequently, when
the headwheel motor starts from rest, the velocity -error signal causes appli-
cation of full power to the motor. The motor then accelerates rapidly toward
the synchronous speed of 480 (or 500) revolutions per second, but as the
tonewheel pulse period approaches the multivibrator period, the velocity-
error signal reduces the power until the speed stabilizes at 240 (or 250)
revolutions per second. At this time, the phase -error detector assumes tight
control and corrects normal variations in headwheel -motor speed.
During manufacture of the headwheel panel, the tonewheel is attached
to the motor shaft at a predetermined angle of rotation with respect to
the headwheel; therefore, a precise phase relationship exists between the
timing of the tonewheel pulses and the positions of the video heads on
the tape. This relationship causes vertical sync to be recorded by video head
1 at a precise spot near the center of the tape width. Slight differences in
the headwheel -to- tonewheel phasing between different machines, or be-

tween different headwheel panels, normally have little effect. When, how-
ever, an electronic splice is made (Chapter 10) , a difference in headwheel-
to-tonewheel phasing between the machine used for the original recording
and the machine used for the splice recording will produce a horizontal
"lurch" in the reproduced picture. Therefore, special splice circuits are
provided in the headwheel -servo and electronic -splice systems to permit
matching the headwheel -to- tonewheel phasing. When the machine is in the
splice -play or splice- record mode, these circuits allow the operator to switch
a variable delay into the path of the vertical reference pulse used for tone-
wheel phase -error detection. A potentiometer in the splice system then
permits matching the phasing by adjusting the delay while observing the
video signal during playback of the original recording. Adjustment of the
potentiometer has the same electrical effect as changing the angle between
the tonewheel and the headwheel.
Minimization of recorded drift or jitter requires fast servo response to
error voltage and therefore high gains in the velocity and phase loops; how-
ever, after a nonsynchronous switch in the incoming video signal. a high
gain in the tonewheel phase loop will cause the headwheel to rephase itself
too rapidly for the capstan servo. Therefore, to permit a smooth transition,
a special circuit designated the Lazy Servo Simulator ( LASS ) temporarily
reduces the phase gain after a nonsynchronous switch ( see Fig. 7 -15) The
.
LASS circuit, actuated only when a nonsynchronous switch occurs during

recording, detects coincidence between the tonewheel pulse and the refer-
ence vertical signal. When actuated, the LASS circuit switches in a poten-
tiometer previously set to provide the required gain reduction and then,
after coincidence is restored, returns the circuit to normal.
Play Mode
During tape playback, the mode of servo operation is selected by a rotary
switch on the tape-sync processor module in conjunction with a push-button
switch on the play control panel. Playback in the tonewheel or Switchlock
mode of servo operation is attained by rotating the switch on the tape -sync
processor module to the TW or SL position, respectively. Playback in the
Linelock or Pixlock mode of servo operation is attained by rotating the
switch on the tape -sync processor module to the normal position and then
selecting the desired mode by means of either the LINELOCK or PIXLOCK
push- button switch on the play panel.
Now before going ahead, let us clarify a rather unfortunate situation in
terminology. When RCA first made synchronous operation available, they
called it "Pixlock." Instruction books referred to this function as either
"Pixlock" or "Linelock" ( since it is in actuality line -to -line lock) , and
either term applied to the same function. For fully synchronous operation
(to be able to use mix and special effects of other local signals with the tape
output) , there must be vertical and horizontal coincidence of the tape out-
put signal with other local signals. At the same time, for playback of a color
tape, the tape system must be locked to local sync for proper control of the
time base. This is necessary even if it is not required to use mix or special
effects with other local sources. The main drawback of fully synchronous
operation is that tape disturbances such as dropouts ( particularly during
the vertical interval) or faulty splices require a complete relocking process
of the servos. Recovery of vertical alignment must come first. Due to the
flywheel ( inertia) effect of the frame rate, this can take from 3 to 4 seconds
in the best of servos. Since the use of local mixing and special effects
involving other local sources with tape playback is limited, another method
of "synchronous lock" has been devised. This method involves locking to
the local sync generator in timing only, dropping the requirement for
exact vertical ( framing) alignment. In the RCA system, the latter mode is
termed "Linelock." Then the fully synchronous mode ( timed and phased
to local vertical sync for use of local mixing) is termed "Pixlock." So the
two terms now cover different functions and are no longer interchangeable.
Tonewheel -In the tonewheel mode of servo operation, the tonewheel
servo loop controls the headwheel motor as in the record mode, except
that either external ( house) sync or a special internal -generator or
power -line signal may be selected for timing the vertical reference pulse.
(The timing signal is obtained by utilizing a switch on the reference
generator module and, if a power -line signal is desired, by changing a
jumper internally.) This mode of servo operation will produce a satis-
factory picture, but tape and reference sync will not necessarily be aligned
because of the absence of feedback from the demodulated (tape) video
signal to the headwheel or capstan servo.
Switchlock-When Switchlock is selected, the headwheel operates in
the tonewheel mode of servo operation, but the capstan servo controls the
tape motion so that the headwheel is on the track containing recorded
vertical sync at the end of a frame when the corresponding vertical inter-
val of reference (local) sync occurs. This coarse framing brings tape and
reference vertical sync close enough to permit roll -free switching be-
tween the recorded output and other video sources, or to permit electronic
splicing; however, it does not correct for errors in placement of the re-
corded vertical sync on the tracks.
NOTE: Before an electronic- splice recording can be made, the servo selec-
tor switch on the tape sync processor module must be in the SL position,
and the machine must be actually playing back tape in the Switchlock
mode of servo operation. If this condition is not met, the splice control
circuit automatically prevents entry into the splice record mode.
As explained in the discussion of the capstan servo system, coarse verti-

cal framing is attained in two stages. In the first stage ( fast lock -up cap-
stan), Switchlock is achieved before a tape playback signal is available
by utilizing the edit pulses recorded on the control track to reset binary
dividers in the capstan servo. In the second stage, which starts after four
seconds from the moment of entry into the play mode, relays substitute
tape frame pulses for the less reliable edit pulses. If no edit pulses were
recorded on the tape, or if they were of poor quality, the capstan servo
would go immediately to the second stage.
Pixlock-In the Pixlock mode of servo operation, the capstan servo
provides coarse vertical framing, as in Switchlock. However, the head-
wheel servo additionally imposes both a tighter phase relationship be-
tween tape and reference vertical sync (tape vertical alignment, or tva)
and a lock between tape and reference (local) horizontal sync (Linelock) .
This combined vertical and horizontal lock permits adjusting the tape
horizontal phasing to obtain coincidence between sync in the video out-
put of the machine and reference sync, so that lap dissolves, special effects,
and similar transitions between the tape signal and other television sig-
nals may be used. In addition, the higher sampling rate used in maintain-
ing Linelock reduces jitter to a low level.
NOTE: Since Pixlock performance depends partly on recorded jitter, best
results in Pixlock playback are obtained with tapes recorded with a low-
jitter servo.
Pixlock operation differs from Switchlock only in the use of error infor-
mation from the Linelock module in addition to the velocity and phase
information provided by the tonewheel servo module. The combined oper-
ation of the capstan and headwheel servos in establishing Pixlock occurs
in two main phases: the tonewheel -tva phase and the Linelock phase.
At the moment the servo mode -selector push- button switch on the play
panel is pressed for Pixlock operation, the machine enters the tonewheel-
tva phase (Fig. 7 -16) In this phase, a variable delay generator (time -
.
modulated vertical circuit) in the Linelock module is switched into the

path of reference vertical sync before the phase comparison circuit in the
tonewheel servo module. A dc control voltage from the tape alignment
(tva) error detector in the Linelock module is fed to the delay generator.
The delay, nominally equal to the period between tonewheel pulses (4166
microseconds on domestic standards or 4000 microseconds on interna-
tional standards) , is modulated by the tva error signal. This signal is a
measurement of the phase difference between tape vertical sync and un-
modulated reference vertical sync.
At the start of the Pixlock operation, the tonewheel servo accelerates the
headwheel motor in the same manner as in Switchlock. While the motor
is accelerating, the tva signal has little effect because either the tape signal
is not yet available or the vertical framing error is beyond the range of
the tva error detector. When, however, the tonewheel velocity loop "locks
in" and the capstan servo achieves coarse framing, the tva voltage assumes
control of the tonewheel servo phase loop. If tape vertical sync occurs late
with respect to reference vertical sync, the tva error decreases the vertical
delay. The time -modulated vertical pulse then occurs earlier than normal
with respect to the tonewheel pulse and produces the same effect as a
later- than -normal tonewheel pulse. The dc tonewheel -servo error signal
therefore becomes more negative. In response to the negative increment,
the headwheel modulator then provides more power to the headwheel
motor, causing an increase in speed. Conversely, if tape vertical sync is
early with respect to reference vertical sync, the delay increases, the tone -
wheel -servo error signal becomes more positive, the motor power de-
creases, and the motor slows down.
The preceding actions rapidly bring the combined tonewheel -tva loop
to an equilibrium condition in which time -modulated reference vertical
sync has the same phasing, with respect to the tonewheel pulse, as the
unmodulated reference vertical sync in the tonewheel mode. The time -
modulated delay is then at a value which brings tape vertical sync very
close to exact coincidence with reference vertical sync. The remaining
error in tape vertical alignment has a fixed and a varying component, both
small (within a few microseconds) because the tva gain is relatively large.
The variable error is due directly to the low sampling rate (60 samples
per second) Further reduction in the fixed error would require an even
.
larger gain not attainable at this low sampling rate. While tape and refer-
Fig. 7 -16. Headwheel servo system in tonewheel -tva phase of Pixlock

or Linelock mode of operation.
ence vertical signals are being brought into alignment, a coincidence de-
tector and lock sense circuit in the Linelock module, independent of the
tva circuit, compare the tape and reference vertical phasing. When verti-
cal coincidence is established, this circuit energizes a bus (lock bus) which
causes relays to switch the headwheel servo to the Pixlock mode.
At the moment the system is switched to the Pixlock mode of operation,
the tva error signal is disconnected, and the Linelock error -detector cir-
cuitry is enabled (refer to Fig. 7 -17). The Linelock error detector com-
pares an output derived from tape horizontal sync at the video demodu-
lator output with a reference derived from local ( house) horizontal sync,
to produce an error signal sampled at the line rate. Filters and amplifiers
separate this signal into a high- frequency, or velocity, ac component and
a low- frequency, or phase, dc component. The high- and low- frequency
error- signal paths have separate antihunt and stabilizing networks. A gain
control in the low- frequency error- signal path permits obtaining optimum
loop gain and phase characteristics. In addition, each output amplifier pro-
vides special dynamic characteristics which compensate for the mechanical
response of the headwheel motor to sudden disturbances.
The low- frequency Linelock error component is fed to the time -modu-
lated vertical circuit in place of the tva error, and the high- frequency
Linelock error component is fed, through a separate path, directly to the
headwheel modulator. When the Linelock error -detector circuit is en-
abled, a time -delay circuit gradually increases the gain to prevent tran-
sients from supplying false information. As the gain increases, the low -
frequency error component assumes control of the time -modulated verti-
cal circuit; thereby it adjusts the delay previously established by the tva
error signal to a value which causes the tonewheel servo to align tape
horizontal with reference horizontal. Simultaneously, the high- frequency
Linelock error signal provides velocity information to the modulator
which locks the frequency of tape horizontal sync to that of reference
horizontal sync.
Completion of all phases of Pixlock operation requires less than five
seconds from the moment of entry into the play mode. When the Line -
lock phase is completed, only the mechanical capabilities of the motor
drive system limit the tightness of control. The high gain and rapid re-
sponse of the combined tonewheel and Linelock system reduce the maxi-
mum phase error between tape and reference sync to less than ±0.07
microsecond for a tape recorded on a machine having the same type of
servo system.
As a distinguishing characteristic of Pixlock operation, the vertical -
coincidence circuit continues to check the phase relationship between tape
and reference vertical sync after the machine enters the Linelock phase.
Therefore, if dropouts or similar disturbances cause loss of vertical coinci-
dence, the lock -sense circuit will temporarily switch the headwheel servo
back to the tonewheel -tva phase until the tva circuit re-establishes coinci-
Fig. 7 -17. Headwheel servo system in lock phase of Pixlock or Linelock.

dence. Resynchronization will normally occur within a time interval of

three seconds.
To permit obtaining the different phase relationships between tape and
reference (local) sync required for operation without ATC or in the vari-
ous ATC modes, delays not previously mentioned are inserted into the
paths of tape and reference horizontal sync before the actual phase com-
parisons are made in the Linelock and tva circuits. The Linelock circuit
itself contains a fixed delay in the reference horizontal -sync path and a
variable delay in the tape horizontal -sync path. Additional delays required
only for ATC operation are inserted by the ATC systems before the ref-
erence signals are fed to the Linelock module.
Linelock-Initially, the Linelock mode of servo operation is similar to
the Pixlock mode of operation. However, in Linelock, once the switch to
the lock phase occurs, both the vertical -coincidence and Switchlock re-
quirements are eliminated. Rapid recovery from tape disturbances such
as dropouts or faulty splices is thus ensured, since the headwheel can im-
mediately lock to any horizontal line without waiting for reframing. The
Linelock mode is primarily useful in normal Color ATC or npc (non -
phased color) playback, where vertical coincidence is not a requirement.
(It should be noted that lack of vertical framing prevents the use of spe-
cial effects.)
When the servo mode -selector switch on the play panel is pressed for
Linelock operation (selector switch on tape sync processor module in Nor-
mal position) and the PLAY push button is pressed, operation of the head-
wheel and capstan servos is, initially, similar to Pixlock operation except
that an eight- second time -delay section of a special control circuit (Pix-
lock Achieved, or Platch, circuit) is energized. If Pixlock is achieved
within the initial eight seconds, the Platch circuit energizes a control bus
one second after the switch to the lock phase occurs (Fig. 7 -18) . The
Play Bus
Platch Bus to
Lock Bus Vertical -Coincidence
From Circuit
Linelock
Circuit Bus Asking For
Linelock Servo
Tonewheel_1.1 TW:8/10 TWi8/10Frame Pulses Platch

Pulses Divider To Capstan -Servo
Normal Trapezoid Generator
Reference Frame Pulses o
o-1 To Capstan -Servo

r-Sample -Pulse Generator
Tape Frame Pulses o t (CT :8/10 Divider)
Fig. 7 -18. Pixlock Achieved (Platch) circuit.

Platch bus then disables the vertical- coincidence circuit of the headwheel
servo, and a relay switches the capstan -servo reference from the local frame
pulse to a frame pulse derived from the tonewheel. However, if Pixlock
is not achieved within eight seconds, the Platch circuit is automatically
energized, and the Platch bus disables the vertical- coincidence circuit. The
lock -sense circuit then switches the system to the lock phase.
The Platch bus thereafter remains energized throughout the playback
period. The Linelock circuitry then functions exactly as described above
under Pixlock, except that a loss of vertical framing does not cause rever-
sion to the tonewheel -tva phase. After a disturbance, the Linelock circuitry,
by rephasing the headwheel, locks the nearest horizontal line of tape video
to reference (local) sync, and the capstan servo, referenced to the tone -
wheel, maintains tracking by adjusting tape motion.
7 -4. THE RCA CAPSTAN SERVO

The purpose of the capstan servo system is to control the speed of the
capstan motor and thus the speed at which the tape runs through the ma-
chine. In addition to the normal tape speed of 15 inches per second
(15.625 in /s on 405/625 -line international standards), the tape may be
pulled through the machine at half -speed (71/2 in /s). The primary ad-
vantage of running the tape at half -speed is that twice as much recording
time can be obtained from a given length of tape as would be obtained
from the same length of tape running at 15 in /s. The increase in record-
ing time results in advantages such as reduction of tape costs, saving on
tape storage space, and reduction in tape distribution expense. Although
the tape speed is nominally 15 or 71/2 in /s, during tape playback it is
necessary that the capstan -motor speed be tightly controlled by the cap-
stan servo system to ensure that exactly four video tracks are pulled past
the rotating headwheel assembly during each rotation of the headwheel.
The capstan servo system in the TR -70 consists of the following mod-
ules in addition to the capstan motor itself:
1. Control -track record /playback (617)
2. Capstan phase (616)
3. Capstan error (615)
4. Capstan oscillator (614)
5. Capstan power amplifier (605)
The following discussion of the capstan servo system presents a brief
overall description of the system operation during tape recording and
playback. In the discussion, all timing references pertain to machine op-
eration on domestic (525 -line) standards. If the machine is operated on
international standards, the appropriate timing differences must be taken
into consideration (i.e., substitute 50 Hz for 60 Hz, 250 Hz for 240 Hz,
etc.).
Also, throughout the discussion of the capstan servo system, the term
"frame pulse" is used frequently. In the tape recording and playback
processes, there are actually several distinct types of pulses occurring at
a frame rate (30 Hz on domestic standards or 25 Hz on international
standards). One type, designated the reference frame pulse, is obtained by
comparing station (reference) horizontal- and vertical-sync signals in a
coincidence gate circuit. During playback, this pulse is used directly as a
reference for the capstan servo. During recording, the pulse is superim-
posed on the control -track signal, and the combined signal is recorded on
the tape as specified by the SMPTE recommended practices. The frame -
pulse portion of this combined control -track signal is referred to either
as an edit pulse or as a control -track frame pulse. During tape editing, this
pulse serves as a frame marker on the tape to permit making roll -free
mechanical splices. Also, during the first four seconds of playback, the edit
pulse is separated from the combined signal picked up by the control -
track head and used in the fast lock-up circuit of the capstan servo.
In addition to the reference and control -track frame pulses, a tape frame
pulse is used during playback in the Switchlock and Pixlock modes of
servo operation. This pulse is generated by feeding tape horizontal and
vertical sync (separated, after demodulation, from the signal picked up by
the video heads) to a coincidence detector. The tape frame pulse, there-
fore, occurs at a definite time during the vertical -blanking interval at the
end of each frame of the tape video signal.
Normal Record Mode

In order to maintain a definite timing relationship between tape
movement and video -head scanning during the recording process, the cap-
stan motor speed is "locked" to that of the headwheel motor. To accom-
plish this, a chain of binary counters is used to divide the frequency of
the 240 -Hz tonewheel pulses by four, and the resulting 60 -Hz signal is
used to drive the single -phase, hysteresis synchronous capstan motor. Be-
cause the tonewheel is physically attached to the headwheel motor shaft,
both the tonewheel and the headwheel rotate at the same speed. There-
fore, since the frequency of the signal driving the capstan motor will fol-
low that of the tonewheel pulses, a "lock" is established between the cap-
stan and headwheel motors. This method of controlling the capstan motor
speed ensures that the tracks laid down on the tape during recording fol-
low a standard geometrical pattern (e.g., that specified by the SMPTE or
CCIR), thus facilitating interchangeability of tapes.
NOTE: In the record mode, the headwheel motor must lock in at its cor-
rect speed before the tonewheel pulse is applied to the first binary counter.
Therefore, until a headwheel lock has been obtained, the counter is driven
by the 240 -Hz local oscillator used during tape playback. This arrange-
ment protects against excessive motor and power-amplifier currents dur-
ing start -up.
The binary counters produce a 60 -Hz square -wave signal which is con-
verted into a sinusoidal signal and fed to a power amplifier. The power
amplifier then provides the power required to drive the capstan motor
(Fig. 7 -19).
To provide a continuous, accurate record of the capstan motor speed
during recording, a 240 -Hz sinusoidal control -track signal is recorded
along the lower edge of the tape by the control -track head located on the
headwheel panel. (The control track is utilized in controlling tape motion
during playback and is analogous to the sprocket holes in motion -picture
film.) The control-track signal is derived from the tonewheel pulse signal,
which passes through a variable delay circuit before being formed into a
sinusoidal waveform, and thus bears a definite phase relationship to the
video heads. (It should be noted that when the recorded control -track
signal is played back, it will not appear to be sinusoidal; because a direct
recording process instead of a linear recording process is used, the signal
will exhibit a characteristic double- humped pattern. This is clarified in
Section 7-5.)
Superimposed on the control -track signal is a stable timing reference
pulse designated the control -track frame pulse. This pulse is produced by
a coincidence gate circuit which compares reference vertical- and refer-
ence horizontal -sync signals. The pulse occurs at a 30 -Hz rate (25 -Hz rate
in international machines) , and thus there will be one control -track frame
pulse for every eight cycles of control -track signal (every ten cycles in
international machines) To meet SMPTE standards with regard to rela-
.
tive phasing between the control -track signal and the control -track frame
pulse, a variable delay is incorporated into the circuit which generates the
control -track signal. Thus, by maintaining a specific phase relationship, the
control -track frame pulses provide physical locations on the tape (approxi-
mately every half inch) at which the tape may be cut and spliced without
danger of picture rollover (loss of vertical sync) during tape playback.
Since the control -track frame pulse may be utilized as a guide for cutting
the tape during tape editing, this pulse is sometimes referred to as an
"edit" pulse.
The level of the combined control-track signal and control -track frame
pulse fed to the control -track head is fairly critical. Therefore, the simul-
taneous playback head is provided as a means of playing back the com-
bined signal immediately after it has been recorded, so that the signal
level may be observed on the cro monitor and corrected if necessary. The
combined signal may also be observed directly on the monitor as it is be-
ing recorded. This allows the operator to position the control -track signal
relative to the control -track frame pulse precisely and to obtain the cor-
rect pulse amplitude. In addition to these presentations on the cro moni-
tor, provision is made for observation of the motor drive signal as a
means of checking the normal output which drives the capstan power am-
plifier during the recording process.
r 7
C N
.0 2N iao
ó V
0
o >
L. N
u--
La, H
N LN ó 0,.°.
ÿ úo .p vN C *C
o {n U o
ti 0_ S
RI
uOiIOW adel
Fig. 7 -19. Capstan servo system in record mode.

Playback Operation
During playback, several modes of operation are possible. These are
described as follows.
Tonewheel-During tape playback in the tonewheel mode of servo
operation, the capstan servo controls the longitudinal tape speed and phas-
ing so that the video heads track properly on the recorded video tracks.
In playback operation, the capstan motor speed is controlled by a local
oscillator rather than by the tonewheel pulse as it is during the recording
process. (See Fig. 7 -20.) The oscillator is "free- running" at a frequency
of 240 Hz; however, if the tape moves too slowly or too rapidly, an error
signal is developed which increases or decreases the oscillator frequency
(and thus the capstan motor speed) to correct the tape speed. The 240 -Hz
output signal from the oscillator is divided by four by the binary counters
and formed into a 60 -Hz capstan -motor drive signal, in exactly the same
manner as is the tonewheel pulse during recording.
Run 9
Edit Pulse
d0 Start
4 -Sec Swi ch
/Control-Track
\RecIPB Head
-r
Con trol -Track
Pulse Generator
(Control Track
RecIPB 6171
111.1
Variable Delay
(Capstan Phase
6161
- CT
:8 Counters
1:10
Inter nation all
(Capstan Phase
6161
Tape Reference Reference Reference Comparator

Vertical Sync Frame -Pulse Trapezoid Bridge
Generator Generator
Reference (Capstan Error (Capstan Error (Capsta Error
Horizontal Sync 6151 6151 6151
Control Voltage)
240 -Hz
Binary
Oscillator
Power Counters
(250-Hz
Capstan Amplifier
605
e4
-- I nternational1
(Capstan (Capstan
Oscillator 6141 Oscillator 6141
Fig. 7 -20. Capstan servo system for playback in tonewheel mode.
The error signal which modifies the local -oscillator frequency to correct
the tape speed is generated from the phase comparison of a pulse derived
from the playback control -track signal and a trapezoid waveform derived
from the reference frame pulse, which in turn is derived from local (refer-
ence) sync. The 240 -Hz control -track signal is played back from the tape
by the control -track record /playback head and is amplified, limited, and
formed into a series of 240 -Hz pulses. The 240 -Hz pulse frequency is
then divided by eight (by ten in international machines) by a binary

counter chain to obtain a series of 30 -Hz pulses (25 -Hz, international) .
These pulses sample the slope of the reference trapezoid waveform to ob-
tain the error signal. Local ( reference) horizontal- and vertical -sync sig-
nals are combined in a coincidence gate circuit to form pulses at the frame
rate, and the reference frame pulses thus formed generate the reference
trapezoid waveform. Therefore, as the tape speed increases or decreases,
the control -track pulse frequency follows this change, and the error signal
generated causes the capstan motor speed to increase or decrease until the
divided frequency of the control -track pulses coincides with that of the
reference frame pulses ( i.e., the capstan motor speed is locked to the ref-
erence pulse frequency) .
During the first four seconds after the PLAY push button is pressed, con-
trol -track frame pulses (edit pulses) if present on the control track, are
,
used to reset the binary counters which divide the frequency of the 240 -
Hz control -track signal. This permits achieving coarse framing (approxi-
mate alignment of tape vertical sync with local vertical sync) before the
vacuum -guide air solenoid is activated and video is obtained, since the
control -track head is already in contact with the tape. Application of the
edit pulses to the binary counters is controlled by the fast lock -up capstan
circuit, which consists of a relay controlled by multivibrators. Although
this circuit is active during the first four seconds of playback regardless
of the mode of servo operation, it is primarily useful in the Switchlock,
Pixlock, and Linelock modes. In the tonewheel mode, after the relay dis-
connects the edit pulses at the end of the four -second interval, the binary
counters are no longer reset. Therefore, if framing is lost because of a
disturbance, it will not be regained, although the frequency- divided con-
trol -track pulses will relock with the reference frame pulses. This is con-
sidered normal for the tonewheel mode, since framing is not a requirement.
Provision is made for continuous manual variation of control -track
phase over a very wide range. Manual phasing is accomplished by varying
the CT PHASE control on the play control panel, which in turn varies a
delay inserted into the control -track pulse path. (The CT PHASE control has
two ranges: a wide range for coarse phasing and a much narrower range
for fine phasing.) The basic purpose of the manual delay is to provide a
means of precisely centering the video heads over the recorded video
tracks. In the tonewheel mode of servo operation, the CT PHASE control
may also be utilized in "slipping tracks" so that video head 1 will play
back information from the video track containing vertical sync (which is
normal operating procedure in the RCA system) or from any of the
three remaining tracks. Provision is also made for manual control of the
capstan motor speed, and thereby of the tape speed, during playback by
interrupting the error signal fed to the 240 -Hz local oscillator and insert-
ing a constant dc current. The constant dc current is inserted by pressing
the red (to speed up the tape) or black (to slow down the tape) CAP-
STAN SPEED button on the play control panel. This feature is useful for
synchronizing two machines which are playing back the same information.
In the Switchlock and Pixlock modes of servo operation, the capstan
servo operates as in the tonewheel mode except that when the edit pulses
are disconnected from the binary counters at the end of the initial four-
second interval, the edit pulses are replaced by tape frame pulses (Fig.
7 -21) . (If no edit pulses are on the tape, or if they are of poor quality,
tape frame pulses are fed to the counters as soon as they are available.)
Use of the tape frame pulses for resetting the counters ensures that the
headwheel will be on the track containing vertical sync at the end of a
recorded frame when the corresponding vertical interval of local sync
occurs. This degree of control, designated coarse framing, or Switchlock,
brings tape and local (reference) vertical sync close enough together to
permit roll -free switching between the tape- machine output and another
video source, or to permit electronic splicing. (In the Pixlock mode of
servo operation, a finer degree of framing, which corrects for errors in
Tape Frame Pulse

30 Hz (25 Hz International)
r
Tape Sync Reset Pulse
Processor 4 -Sec Switch Gate
Headwheel
622
(Part of
Video Proc) Edit Pulse
c
- (Capstan Phase
6161
Reset Pulse
1
Coincidence Sens
CT
Control -Track
Variable Delay 8 Counters
Pulse Generator 1:10
Control -Track
RecIPB Head International)
(Control Track (Capstan Phase
( Capsta Phase
Rec/PB 617) 6161
6161
Tape
Reference Reference Reference Comp rator
Vertical Sync Frame -Pulse Trapezoid Bridge
Generator Generator
Reference (Capstan Error (Capstan Error (Capsta Error

Horizontal Sync 615) 615) 615)
Control Voltage
240 -Hz
Binary
Oscillator
Power Counters
1250 -Hz
Capstan Amplifier 4
International)
605 (Capstan (Capstan
Oscillator 614) Oscillator 614)
Fig. 7 21. Capstan servo system for playback in Switchlock or Pixlock mode.
placement of vertical sync on the tracks, is achieved by utilizing the tape

vertical alignment [tva] circuit in the headwheel servo.)
Eight 240 -Hz control -track pulses (or ten 250 -Hz pulses on 405/625 -
line international standards) are fed to the binary counters per frame,
but only a particular one of these eight (or ten) pulses occurs when the
headwheel is on the track containing vertical sync at the end of a frame.
Use of the reset pulse ensures that the single output pulse provided by the
binary counters during each frame coincides with this particular control-
track pulse. If no reset pulse were applied (as in the tonewheel mode after
the initial four seconds), there would be only one chance in eight (or
ten) of obtaining this condition, since the counter output pulse could
coincide with one of the other control -track pulses occurring during the
frame period. Consequently, there would be only once change in eight (or
ten) of achieving coarse framing (Switchlock) .
The tape frame pulse utilized in resetting the counters is developed

from tape vertical and tape horizontal signals in a coincidence gate cir-
cuit. If the tape and reference frame pulses are aligned when the machine
enters the Switchlock mode of servo operation, the reset pulse will have
no effect on the binary counter circuits. However, if the frame pulses are
not aligned, the reset pulse will reset the binary counters and thus cause
the machine to "shift tracks" until sync alignment is attained.
During tape playback in the Switchlock or Pixlock mode of servo op-
eration, the CT PHASE control may be utilized for precise centering of the
video heads over the recorded video tracks as it is during playback in the
tonewheel mode. In the Switchlock or Pixlock mode, it is not possible to
use the CT PHASE control to "slip tracks" as it is during tonewheel opera-
tion; however, the need for "slipping tracks" is eliminated by the fact that
it is possible to "shift heads" by utilizing the HEAD SELECT switch on the
reference generator module in the headwheel servo system.
Linelock -The initial phases of Linelock operation are identical to those
of Pixlock operation, but once the switch to the final, or "lock," phase
occurs in the headwheel servo system, the framing requirement is dropped
by (1) disconnecting the tape frame pulses from the binary counters in
the capstan servo system and (2) switching the reference for the capstan
trapezoid from reference frame pulses to pulses derived by dividing the
frequency of the tonewheel pulses by eight (or ten on international
standards) . Refer to Fig. 7 -22. This action leaves the capstan free to
follow changes in headwheel phasing (which occur after disturbances),
thereby maintaining the tracking adjustment made with the CT PHASE
control. The changes in headwheel phasing occur after a disturbance that
causes loss of Pixlock, because the Linelock circuits in the headwheel servo
system lock the tape horizontal sync to a noncorresponding line of local
(reference) horizontal sync.
The tonewheel divide -by-eight pulses are obtained from binary counter
circuits in the tonewheel servo module. These counters, which are similar
Magnetic
240 Hz
HeadwherI
¡ Tonewheel i 218 +8 Pulse
(250 Hz s8 Counters
i -
'
Tonewheel Servo
International) Processed Tonewheel Pulse UID - I Linelock
International)
(Tonewheel (Tonewheel i Norm
Processor 6121 Servo 6111 i
Video Tape Frame Pulse

30 Hz (25 Hz International)
Reset
1I
Linelock 4-Sec Switch
Reset -Pulse Reference
Video System ° Ruk n
Gate Frame -Pulse
Generator
]
_ .,_} Start
Headwheel (Tape Sync Norm (Capstan Phasa (Capstan Error
Processor 622) 6161 6151
Edit Pulse Reset P Ise {

Coincidence Sense
CT Reference
i
Reference
Control-Track
Control -Track
RevIPB Head
Pulse Generator
(Control Track
-r
Variable Delay
(Capstan Phase
- s 8 Counters
1:10
In ter n utionall
(Capstan Phase
Horizontal
Sync
Vertical
Sync
RecIPB 617) 6161

6161
Divided Coot d Track)
Tape Comparator Reference

Bridge Trapezoid
Generator
(Capstan Error (Capstan Error

6151 615)
f
Control Voltage) Were ce Trapezoid
240-Hz
f-
Binary
Oscillator
Power Counters (250 -Hz
Capstan Amplifier a4
International)
605 (Capstan (Capstan
Oscillator 6141 Oscillata 614)
Fig. 7 -22. Capstan servo system for playback in Linelock mode.
to the control -track pulse counters in the capstan servo system, are reset
by the reference frame pulses so that the divide -by -eight pulse coincides
with the tonewheel pulse closest to the reference frame pulse. This elimi-
nates picture disturbances during the transition period after the capstan -
servo reference is switched from reference frame pulses to tonewheel
divide -by -eight pulses.
When the headwheel phasing changes, the tonewheel -pulse counting
speed will vary depending on the direction of change, but the counter
output will remain properly phased with respect to the reference frame
pulse. If the headwheel speed increases, the counters will count tonewheel
pulses faster, and the capstan motor speed will increase until the frequency -
divided control -track pulses are again locked to the frequency- divided tone -
wheel pulses.
Splice Modes -In
machines equipped with the accessory electronic splic-
ing feature (Chapter 10) , the capstan servo system operates as previously
described except when an insert or add -on splice recording is being made.
(In an insert splice recording, the video on a section of a previously re-
corded tape is erased and new material is recorded in its place, leaving
portions of the originally recorded video immediately preceding and fol-
lowing the newly recorded section. In an add -on splice recording, however,
the new material is added at the end of the original recording.) As far as
?2 TELEVISION BROADCASTING
the capstan servo system is concerned, the essential difference between the
insert and add -on mode of splice operation is as follows: In the insert
mode, the capstan servo system operates as during normal tape playback,
and the previously recorded control track is not erased but is allowed to
remain on the tape. In the add -on mode, the capstan servo system operates
as during normal tape recording, and the original control track (if any ) is
erased so that a new control track may be recorded.
In either type of splice recording, the capstan servo must be controlled
so that no sudden change in tape speed occurs at the junction of the old
and new recordings, and the tape speed during recording will be the same
as during playback. These results cannot be obtained by driving the cap-
stan motor with a signal derived from the frequency- divided tonewheel
pulses, as in normal recording, since the tonewheel pulses provide no indi-
cation of the actual tape speed present when the old recording was made.
Therefore, during a splice recording, the capstan is driven, as in playback,
with a signal obtained by dividing the frequency of the output from the
240 -Hz oscillator. The method of controlling the oscillator frequency de-
pends on whether an insert or add -on splice recording is being made.
For an insert recording, the control -track signal previously recorded (at
the same time as the old video signal) is already present along the entire
length of tape, and this signal contains precisely the desired information
regarding the speed of the old recording. Therefore, instead of erasing the
old control track and recording a new one as the new video information
is being recorded, the old control track is retained and used, as in Switch -
lock operation, to derive an error signal which controls the frequency of
the oscillator. Operation of the entire capstan servo system during insert
splice recording is actually identical to Switchlock playback except that
the reset -pulse gating circuits for the control -track binary dividers are dis-
abled so that the binary counters cannot be reset by noise pulses during
the record interval of the electronic splicing process.
In an add -on splice recording, the section of tape on which the new
video recording is to be made may not already contain a recorded control -
track signal, since this section begins at the end of the old recording.
Therefore, a new control track must be recorded, as in normal record
operation. During the transition from playback to recording, the fre-
quency of the capstan oscillator ( and thus the capstan motor speed) is
controlled by an adjustable dc voltage from a capstan memory circuit, in-
stead of by an error signal. The dc voltage is related to the original re-
cording speed by adjusting a potentiometer on the splice logic module
while playing back the originally recorded section of the tape before start-
ing the add -on recording. For testing the potentiometer adjustment, a
momentary -contact toggle switch on the splice logic module controls a
relay which permits opening the capstan servo loop and temporarily dele-
gating control of the oscillator to the capstan memory circuit while the
machine is in the splice play mode.
NOTE: Before a splice recording is made, the machine must be playing

back tape in the splice play mode, the selector switch on the tape sync
processor module must be in the SL position, and Switchlock operation
must actually be achieved.
7 -5. THE CONTROL -TRACK RECORD AND

SIMULPLAY WAVEFORMS
The great difference between the control -track record and playback
waveforms was mentioned in the preceding section. This difference will
be examined in more detail now.
Note from Fig. 7 -23A that the control -track head is adjacent to the head -
wheel (it is actually mounted on the vacuum guide) and that farther
down the line is a stack containing simulplay heads. This stack includes a
control -track monitoring head to show that a control track is actually
being recorded while the machine is in the record mode. In the record
mode, either the control -track record waveform ( waveform B in Fig.
7 -23B) or the control -track playback waveform (C) can be monitored. The
question always arises as to why this playback waveform results from a
sine -wave recording signal.
Before we continue, let us emphasize a point. In the playback mode, the
same control -track head that made the control -track recording is being
used. When you monitor control-track playback in the playback mode,
you are seeing the signal as it comes from the regular control -track head
on the vacuum guide. This signal also looks like waveform C. But, the
only time you see a signal from the simulplay control -track head is in con-
trol -track playback monitoring during the recording mode.
The proper amplitude of control -track signal in the recording head is
that which drives the tape to the verge of saturation. Therefore, while the
system is recording, the control -track record amplitude should be brought
up (while the simulplay head signal is being observed) until the "shoul-
der" appears, indicating the verge of saturation. See waveform C in
Fig. 7 -23B.
Now let us see what is actually happening. The sinusoidal record cur-
rent (waveform D) is applied to the control -track head. The transfer
characteristic of magnetic tape is nonlinear, as shown by the 13,-Ir. curve
(E). The flux density recorded is shown at F. Note that the zero axis of
the recording current (waveform D) , since it corresponds to the most
rapidly changing area, results in the maximum recorded flux change of
waveform F. The peaks of waveform F correspond to maximum flux of
the indicated polarity (north or south) .
Now refer to the playback voltage waveform (C) The region of maxi-
.
mum recorded flux of waveform F, which is on the verge of saturation,

results in the zero axis of the playback voltage. The "shoulder" of wave-
form C shows the verge of saturation on the peaks of waveform F.
Control -Track Head Sim ulplay Heads

Dr um (Rec /Play) (Includes Control Track)
Tape
Guide
From Control -Track

Play Amp
Play Rec or Play Simulplay Monitoring During
Rec Record Mode Only
From Control -Track I
Record Amp
CR0 Feed
(A) Head functions.
Edit Pulse In Playb ck

Control -Track
Playback Signal
Sine Wave
Shoulder
Rec Edit
Spurious
Response
Edit Pulse
(Every 8th Wave)
Record Signal
Ir
-L
Max Record Flux Change
(B) Waveforms.
Fig. 7 -23. Control -track record and playback waveforms.
The points on waveform C corresponding to maximum flux change in

waveform F are points X and X'. The "hump" between these points is
the result of the nonlinear transfer characteristic (curve E) of the tape.
The edit pulse, since it is made to occur at maximum flux, appears at the
zero voltage axis of waveform C.
Notice the "spurious pulse" occurring at point A on waveform C. (In
some systems, this pulse might be located immediately to the left of the
larger -amplitude playback pulse.) This is nothing but cross talk between
the actual recording head and the simulplay head. It occurs only during
the record mode. Any noticeable drift in this spacing during the record
mode is an excellent indication of jitter that occurs during the recording
process.
Please bear in mind that the audio and control -track simulplay heads
are only indicators that the signals are being recorded on the tape. In nor-
mal playback operation, the control -track and audio recording heads are
used to play back the signals, so that proper timing is maintained. There-
fore, these heads are properly termed "record- play" heads.
7 -6. THERCA HEADWHEEL UNLOCK

(HW UNLOCK) INDICATOR
Another function in need of some explanation is the warning indicator
on the RCA TR -70 designated as the HW UNLOCK light.
NOTE: It is absolutely imperative that the reader have a background in
logic circuitry at least equivalent to Chapter 3 of Television Broadcasting:
Systems Maintenance, by Harold E. Ennes (Indianapolis: Howard W.
Sams & Co., Inc., 1972).
The HW UNLOCK indicator alerts the operator whenever the headwheel

drops out of the lock condition. It is driven by the HW UNLOCK indicator
circuit (Fig. 7 -24) and warns that not all the conditions required to sat-
isfy a particular servo mode have been met. The lamp circuit is operative
only when the run bus (pin 18) is activated (grounded) This occurs in
.
any of the modes in which the headwheel is running, i.e., during record,
set-up, wind, standby, and any one of the playback modes.
When the machine is first turned on, the circuit is inoperative. Placing
the machine in any of the above run modes causes the HW UNLOCK lamp
to light, indicating that the headwheel has not yet come up to the proper
speed. Once the headwheel achieves correct speed and is locked in, the
indicator lamp will go out. Should the headwheel go out of lock for more
than 0.1 second, the HW UNLOCK lamp will light again.
A logic block diagram of the Hw UNLOCK indicator circuit is shown in
Fig. 7 -24. It illustrates the conditions required to extinguish the light.
Wherever AND and OR gates are shown, a small circle at the input or out-
put means that a negative (or low) input or output will satisfy the con-
ti
-8-
2Ó
Q
3
X
ÿU -O.
cro U
>
CC o
H
o óá
Ea
s
1
,á
wa
Eg
6
oc4 oo. oe, ó o7.1 c. o

5 É - m
ám z
z
y Ú
Fig. 7 -24. Logic block diagram of HW UNLOCK indicator circuit.

-
ditions of the gate. Lack of a small circle indicates that a positive (high
ground potential, usually ) input or output level will satisfy the conditions
of the gate.
Operation of the basic circuit is as follows: Tonewheel (TW) pulses
from the tonewheel processor module, and tonewheel phase -sample pulses
from the reference generator module are fed through emitter followers
Q22 and Q23 to transistors Q19 and Q21. These transistors form a coin-
cidence gate and are normally conducting heavily. With the arrival of a
negative -going TW pulse and a negative -going phase - sample pulse in
coincidence, the gate is opened, and a positive -going output pulse is ob-
tained at the collectors. This pulse, differentiated and clipped, is fed as a
trigger to monostable multivibrator Q17 -Q18. The positive -going square
wave at the collector of Q17 is applied to an integrator emitter follower
stage, Q16. The resulting integrated waveform fed to the base of lamp
driver Q15 is sufficient to cut off Q15 (which is normally saturated) and
prevent the HW UNLOCK indicator lamp from lighting.
A loss of coincidence between the TW pulses and the phase -sample
pulses would cause no trigger to be fed to the base of transistor Q18, and
no square wave would be fed from the multivibrator. The charge on ca-
pacitor C15 would fall (go more negative), transistor Q16 would con-
duct, providing bias current for Q15, and the light would be switched on.
However, transistor Q17 also functions as an AND gate and Q15 as an OR
gate. A discussion of the conditions necessary to fulfill the requirements
of these gates is given below.
As mentioned above, the logic block diagram illustrates the conditions
required to extinguish the light. In the stop mode, it will be extinguished
because the run -bus input to the Q15 OR gate will be low. It will be out
in any run mode because the other input to the Q15 OR gate will go high
when the inputs to the Q17 AND gate have been satisfied by the appro-
priate lock conditions.
In all run modes (with one exception to be described later) , coinci-
dence of the TW pulse and the phase- sample pulse is required to provide
a high level to input 1 of the Q17 AND gate. Coincidence of these pulses
is indicative of tonewheel lock and is a necessary, but not sufficient, con-
dition to assure that all the requirements of a particular servo mode have
been met. Input 2 to the Q17 AND gate supplies the remaining condi-
tions, involving somewhat more complex logic, as shown in the diagram.
These conditions are described by modes below.
Record, Set-up, Standby, and Wind Modes

In the record, set -up, standby, and wind modes, input 2 to the Q17 AND
gate is high because the play -bus input to OR gate A is low. This produces
a high output to OR gate B, which in turn produces a high output. Essen-
tially, a tonewheel lock is the only requirement to extinguish the light
in these modes.
22S TELEVISION BROADCASTING
Tonewheel Playback Mode

If the tonewheel mode of servo operation has been selected in playback,
the Switchlock (SL) bus and the bus asking forLinelock /Pixlock servo
(BALPS) inputs to AND gate C will both be low. Therefore, the output
from AND gate C to OR gate A will be low, producing a high output to OR
gate B, which in turn produces a high output to AND gate Q17. In this
mode also, tonewheel lock is the only requirement to extinguish the light.
Switchlock Playback Mode

If Switchlock has been selected as the playback mode, the requirements
for lock are dependent on whether the Electronic Splicing Accessory is in-
stalled in the machine and whether a splice mode has been selected.
Switchlock With Splice Modules Not in Machine -In this mode, the SL
bus input to AND gate D will be high, and the SL coincidence signal input
to AND gate D will be an open circuit (equivalent to low) . Therefore, the
output of AND gate D will be high, and the output of OR gate B to AND
gate Q17 will be high. Tonewheel lock is the only requirement to ex-
tinguish the light.
Switchlock with Splice, Normal Insert or Add -On Modes With Fixed
TW Selected -When the splice modules are in the machine, a coincidence
detector in the splice logic module generates a signal which indicates when
the reference vertical and tape vertical pulses are coincident. Therefore,
although the SL bus is high, the SL coincidence signal input to AND gate
D will not go low until coincidence has been achieved. This will nor-
mally occur after tonewheel lock has already been achieved. Consequently,
tonewheel lock and vertical coincidence are required to extinguish the
light.
Switchlock with Splice, Add-On or Insert Mode With Variable TW Se-
lected -When one of the splice modes has been selected, delays are in-
serted in the reference and signal paths of the tonewheel servo to allow
a controllable adjustment of the phase of the servo relative to the refer-
ence. This has the effect of moving the phase of the tonewheel -phase
sample pulse relative to the undelayed tonewheel pulse so that they are
not coincident when a headwheel lock has been achieved. Therefore, for
this mode only, the necessity for coincidence is removed by energizing the
special splice bus, which produces a high input to OR gate Q20. This will
produce a high input to AND gate Q17 as long as the tonewheel -phase
sample pulses are present, regardless of any coincidence. The other input
to AND gate Q17 still requires vertical coincidence, as described above, to
extinguish the light.
Pixlock Playback Mode
The Pixlock playback mode requires that the lock-bus input to OR gate
B go high, which produces a high output to input 2 of AND gate Q17. The
logic driving the lock bus is contained within the servo modules and
requires that the Pixlock playback mode be selected and vertical coinci-
dence be achieved.
Linelock Playback Mode

In the Linelock playback mode, the requirements for extinguishing the
light are the same as for the Pixlock mode; i.e., the lock bus goes high and
tonewheel coincidence is achieved. However, the logic within the servo
modules which drive the lock bus is changed. During initial start -up of
the machine, operation is identical to the Pixlock mode. When vertical
coincidence is achieved, the lock bus goes high. However, should vertical
coincidence not be achieved for some reason, the lock bus still goes high,
after an 8- second delay. After initial lock is obtained, whether by achiev-
ing vertical coincidence or by the 8- second delay, the lock bus will remain
high until the machine enters the stop mode. Thereafter, headwheel un-
lock will be indicated only by the loss of tonewheel coincidence.
Lazy Simulated Servo (LASS Control)

The lazy simulated servo (LASS) control circuit, consisting of tran-
sistors Qil, Q12, and Q32, is initiated by the HW UNLOCK indicator cir-
cuit and is operative only in the record mode. Under normal conditions
during the record mode, the HW UNLOCK indicator output will have no
effect. However, when a loss of coincidence between tonewheel pulses and
tonewheel -phase sample pulses occurs during any of the run modes, tran-
sistor Q15 is switched on and its collector rises toward ground. Normally,
transistor Q32 is biased on, and this change in potential is coupled out
of the indicator module to the tonewheel servo module where it activates
a relay, reducing the phase gain of the tonewheel servo during the inter-
val of headwheel unlock, in record only.
When the machine is initially placed in the record mode, ground level
appears at pin 6, and consequently at the emitter of Q11 and at capacitors
C12 and C13. While the headwheel motor is coming up to normal speed,
transistor Q11 is held off (approximately 4 seconds) by the positive level
change coupled through capacitors C12 and C13. Transistor Q12 is then
biased on, biasing Q32 off and preventing a level change from appearing
at pin 4. After approximately 4 seconds, transistor Ql1 turns on, biasing
Q12 off and Q32 on, thereby allowing a level change to be coupled out
of pin 4.
7 -7. THE VACUUM -GUIDE SERVO
The function of the vacuum -guide servo system (formerly termed

"shoe" servo) , shown in simplified block-diagram form in Fig. 7 -25, is to
control the position of the vacuum guide with respect to the rotating head -
wheel. The position of the guide is one of the factors determining the
compatibility of playback between tapes recorded by video heads that

have different degrees of wear from one machine to another. Thus the
operation of the guide servo in maintaining the proper guide position
during tape recording and playback is of great importance.
Essentially, the guide servo system consists of the guide servo module
(No. 506) , the guide actuating mechanism (located behind the head -
wheel panel) , and the manually operated guide positioning control uti-
lized optionally during tape playback (located on the play panel) . The
guide servo module contains the electronic circuitry which controls the
Sync
Video
Separator
Playback
(Module 622) System , Video Headwheel
Senses 960 -Hz

Timing Variations
Tape
Motion
ti
Guide
/
Guide Position Determines
Tape Stretch
Er or
Detector Amplifier Guide
Potentiometer
Drive Motor
(Module 6211 (Module 5061
Fig. 7 -25. Simplified diagram of guide servo system.
guide servo system. Operational controls located on the module front

panel include the guide positioning control, which may be utilized prior
to recording; the record control switch, which allows selection of a fixed
or manually preset guide-position reference during recording; and the
playback control switch, which allows selection of manual or automatic
guide positioning during tape playback. The guide actuating mechanism
consists of a two-phase motor coupled through reduction gears to the
headwheel panel and the vacuum guide.
Fig. 7 -26 shows a cross -sectional view of the vacuum guide and head-
wheel, indicating tape stretch due to indentation caused by the video
magnetic heads. As the headwheel rotates, the video magnetic heads (pro-
truding beyond the headwheel rim by approximately 1.0 to 3.2 mils, de-
pending on video -head wear) cause the tape to stretch in a localized in-
dentation which travels across the moving tape. This stretching of the
tape is essential to correct for minor dimensional variations resulting from
video-head wear, differences in machines, and minute expansion or con-
traction of the tape. In recording, the guide position is normally fixed at
a calibrated setting for the particular headwheel panel in use. This ensures
the application of a constant pressure between the tape and the video
heads, thus resulting in a fixed degree of tape stretch throughout each re-
Tape
wide
07 Atmospheric Air
Pressure Holds
to Tape Against Guide
Fig. 7 -26. Cross section of vacuum

guide and headwheel. Guide Pole \`
Piece
Tape Motion
Courtesy RCA
cording. (The vacuum -guide position may be preset prior to recording,

by means of the RECORD GUIDE POSITION control on the front panel of the
guide servo module, if other than standard guide pressure is desired.)
NOTE: The SMPTE test tape (MI -40793 for 525 -line standards, Mí-
40797 for 625 -line standards) is used for comparison in playback to de-
termine the proper tape stretch.
To attain optimum machine performance during tape playback, the

relative head -to-tape speed must be held constant. The factors which de-
termine the relative head-to -tape speed are the angular velocity" of the
headwheel and the position of the vacuum guide. The angular velocity of
the headwheel is maintained at a constant value by the headwheel servo sys-
tem; thus, angular timing errors are eliminated. However, unless the
vacuum-guide position is tightly controlled, timing errors which cause
certain geometric distortions in the picture will occur.
One form of geometric distortion is caused by time -displacement errors
which appear in the form of discontinuities at the instant of video -head
switching. This form of distortion occurs when the "parallel alignment"
of the guide center and headwheel axis is incorrect ( i.e., when the pres-
sure of the video heads against the tape is insufficient or excessive) Dis- .
tortion of this kind is designated "jogs" or "skew," and is shown in Fig.

7 -27. Insufficient head -to -tape pressure causes jogs having a positive slope,
whereas excessive head -to -tape pressure causes jogs having a negative
slope. If a pulse train containing the timing discontinuities is applied to
a horizontal oscillator having a "flywheel" time constant slightly greater
than the 16 -line head -scanning interval, a uniform train of average -
frequency pulses is produced in which the timing variations are elimi-
nated (Fig. 7 -28). The synchroguide and similar circuits in television
receivers, and hence the horizontal- deflection circuits, respond similarly.
However, timing errors in blanking and picture information remain, and

these errors cause the jogs mentioned above. The function of the guide
servo system during playback, therefore, is jog correction. This correction
may be accomplished manually or automatically, depending on the se-
lected conditions for the machine.
Fig. 7 -27. Effect of 1 -mil guide
error on television raster.
Another form of geometric distortion which may appear in the picture

occurs when the "perpendicular alignment" of the guide center and head -
wheel axis is incorrect ( i.e., when the head-to -tape pressure varies as a
video head scans the moving tape) . This form of geometric distortion is
designated "scalloping," and its correction is entirely mechanical. (A third
form of geometric distortion, appearing in the picture as steps, is due to
IIIIIIIJIIIIIIII
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 1
A
Horiz -Sync Pulses
Produced by Head 1
I10I
2
I
3 4 5
Horiz -Sync Pulses
Produced by Head 2
11111111 IIIIIIIIIIIII o Average - Frequency

Pulses Produced by
Tape Sync Processor
OEnlarged Views of
i
Portions of (A)
Timing of pulses
coincides with
II
I
I
L
i
Time D'scontinuity Resulting From
Inadequate Tip Protrusion (Shoe Too Far Out)
average near
OEnlarged Views of
center of head
scanning 114=1. Portions of (B)
interval. II
Enlarged Views of
F Portions of ICI
Fig. 7 -28. Timing errors resulting from shoe position errors.

slight errors in the physical placement of the video heads around the cir-
cumference of the headwheel.)
Movement of the vacuum guide is controlled by a mechanical linkage
from a two -phase induction motor. In the play or record mode, the refer-
ence winding of the motor is continuously excited by the power -line volt-
age, and the control winding is driven by an error signal. Therefore, the
motor shaft normally remains stationary, and it develops torque in one
direction or the other only in response to an error signal. ( The direction
of the torque, and thus the direction of the motor -shaft rotation, is de-
termined by the phase relationship between the 50- or 60 -Hz voltages ap-
plied to the reference and control windings.) The error signal may be
developed manually (during tape playback or prior to recording) or
automatically (during playback only) If the guide position is controlled
.
manually, the error signal is developed when a voltage unbalance exists

between the center arms of two potentiometers. One of the potentiometers
is geared to the guide -motor shaft. During playback, the second poten-
tiometer is the GUIDE POS control on the play panel; prior to recording,
the second potentiometer is the RECORD GUIDE POSITION control on the
guide module. If the guide position is controlled automatically (possible
during playback only) , the error signal is developed in the video and
sync processing system whenever the tape horizontal -sync frequency varies
from the standard line frequency (15,750 Hz in machines operating on
525 -line standards with a 60 -Hz field rate).
When guide -motor torque is developed, it causes a lead screw to move
in or out (depending on the direction of the torque) The lead screw acts
.
as a stop against which the guide control arm rests when the guide solenoid
is activated. The extension of the lead screw therefore determines the posi-
tion of the vacuum guide and thus the pressure exerted on the tape by the
video heads. The guide solenoid is actually a pneumatically operated pis-
ton activated by an air valve which is in turn controlled by a relay. The
air -valve relay is driven by a signal from the vacuum -guide diode matrix
board. The vacuum -guide solenoid logic circuit requires that both a head-
wheel velocity lock and a play or record command be present before the
solenoid driver permits the guide to engage. Therefore, the vacuum guide
will engage only when the play or record push button has been pressed
and the headwheel motor has come up to speed (as evidenced by the
velocity -lock -sense control -bus potential developed in the tonewheel servo
module) Thus, it may be seen that if the machine is operated in the set -up,
.
standby, or wind mode, or if the headwheel motor fails to lock in properly

during tape playback or recording, the guide solenoid is deactivated, and
the vacuum guide is held away from the tape to prevent possible damage
to the headwheel and magnetic heads.
An additional precaution is taken to guard against headwheel and
magnetic -head damage when tape is being played back with the guide
position controlled automatically. This precaution consists of circuitry in
the internal reference module which detects an afc lock. The output from
this circuit controls a relay which provides a ground potential when the
afc lock is attained. The ground potential, fed to the guide servo module,
allows the guide to operate normally in the automatic mode. However, if
an afc lock has not been attained or is lost during playback, ground is re-
moved, and the guide automatically reverts to a manually preset safe posi-
tion. This precaution is important, since an erratic error signal may cause
excessive tape pressure against the headwheel.
To prevent the guide from "hunting," or oscillating, during normal
(nonsplice) recording, the guide motor is disabled following an initial
delay of several seconds that allows the guide to reach its proper position.
If the machine is equipped for electronic splicing, during the splicing
operation the guide motor is disabled immediately when the RECORD but-
ton is pressed; thus, there will be no change in guide position when the
switch from splice play to splice record is made.
In both the Ampex and RCA vacuum -guide servo systems, the circuits
represent an adaptation of a standard servo system designed for a variety
of positioning applications. The main output device for the system is a
motor that normally remains stationary and moves only slightly in one
direction or the other in response to error signals. In general, the motor
automatically seeks to position itself so as to keep the error signal at zero.
The error -detector circuit tuned to 960 Hz (Fig. 7 -25) is used to sense
the presence of timing variations at the head -switching rate. The polarity
of the 960 -Hz signal developed in this circuit indicates the direction of
the timing errors. In the circuitry of the guide -servo chassis, the 960 -Hz
signal is amplified, clipped, and rectified. To preserve the direction sense
in the error signal, the rectifier is a bidirectional type, utilizing 960 -Hz
clamping pulses derived from the tonewheel attached to the headwheel
motor. The clipping is severe enough that if there is any significant error
signal at all, the rectifier produces a fixed dc level of a polarity indicating
60 -Hz Chopper
Motor Winding
60 Hz
117 -Volt Line
+24 V
Guide -Position 24,

Amp
Control Motor
-24 V
Follow Potentiometer
(Rotated by Motor Shaft)
Fig. 7 -29. Basic operation of tape -guide servo.

the direction of the error. The rectifier error signal is converted to 60 -Hz
ac by the action of a chopper -type modulator (Fig. 7 -29) and then ampli-
fied to the power level necessary to drive the guide- positioning motor.
The motor is a two -phase type in which one phase is continuously ex-
cited by the power line, while the other is driven by the amplifier. The
motor develops torque only when a signal is delivered by the amplifier;
the direction of this torque depends on the polarity of the amplifier signal.
The motor then drives the guide through an appropriate mechanical
coupling. The motor movement is limited to only a few revolutions, and
the maximum displacement required for the guide is only about 5 mils.
The servo loop is completed through the video heads, slip rings, video -
playback system, and a sync separator (Fig. 7 -25) .
A much smaller servo loop is used for manual positioning of the guide
during recording operations and during special cases of playback opera-
tion. For manual operation, a potentiometer on the motor shaft is com-
pared with the setting of a second potentiometer, which is mounted on
the control panel (Fig. 7 -29) The motor turns until the voltage at the
.
arm of the motor potentiometer (follower) is the same as that on the

manual guide -position control, resulting in zero signal at the output of
the chopper.
EXERCISES
Q7 -1. Does the same head which records the control -track signal play back
this signal in the playback mode?
Q7 -2. How is the proper amplitude of the control -track recording signal
determined?
Q7 -3. Name the simulplay heads and give their functions.
Q7 -4. How do you control which head records the vertical -sync interval?
Q7 -5. How do you control which head plays back the vertical -sync interval?
Q7 -6. Give the sequence of headwheel (drum) lock-up in going to the fully
synchronous playback mode of operation.
Q7 -7. (A) Give the sequence of capstan lock -up in going to the fully syn-
chronous playback mode of operation. (B) What requirement for a
synchronous operating mode is dropped when phasing is not required
for special effects or mixing? (C) Why must the capstan reference be
changed when going from A to B?
Q7 -8. Define: (A) Edit pulse. (B) Control -track frame pulse, (C) Refer-
ence frame pulse. (D) Tape frame pulse.
Q7 -9. When it is not required to mix other local signals with the tape signal
output, is it necessary to have the servos under control of the local sync
generator for color tapes?
Q7 -10. In most recent tape systems, why is a trapezoidal waveform generally
used instead of a sawtooth waveform?
CHAPTER 8
The Tape -Dropout Compensation

System
The Color Dropout Compensator (DOC) eliminates or greatly reduces

the effects of dropouts in tape- recorded television signals. A dropout is a
brief reduction in the rf level of the reproduced fm carrier due to irregu-
larities in the magnetic tape surface or due to dust or other foreign matter
on the tape or magnetic heads. On the monitor, the loss appears as a dis-
tracting streak. Multiple dropouts can severely degrade the picture. The
Color Dropout Compensator prevents such effects by replacing the miss-
ing information with stored luminance and color information from the
previous scan line.
The reproduced video signal is continuously stored in the DOC by a
delay system having a delay time equal to the duration of one scan line.
So long as there are no dropouts, the recorder output is the nondelayed,
direct video. If the reproduced rf carrier drops below a preset level, the
delay system supplies the video signal, substituting information from the
previous scan line for the missing information. Because of the similarity
between successive scan lines, the viewer will not be aware of the sub-
stituted signal.
The delay system supplies the stored video on demand through a fast -
acting switch circuit. A sensing circuit continuously monitors the repro-
duced rf and actuates the switch circuit whenever a dropout occurs. Since
the detector is sensitive to loss of rf rather than video, the circuit is not
actuated by "recorded -in" video dropouts. These appear when a duplicate
is made from an original tape containing dropouts, unless a DOC was
used during the dubbing process to remove them.
The difference between the Color DOC and a monochrome DOC is that
the Color DOC correctly replaces the color information as well as the
luminance information. If the complete video signal (including the color
information) were delayed exactly one scan line and then inserted as sub-
stitution material for a dropout, the color signal would produce the wrong
236
THE TAPE -DROPOUT COMPENSATION SYSTEM 237
color as viewed on the monitor. This is because the color -subcarrier signal
frequency (3.579545 MHz) is exactly 2271/2 times the horizontal line
rate. Therefore, the color information is changed in phase 180° on each
successive horizontal scan line. This makes it necessary to separate the
luminance and color information, because the color signal must be phase
inverted to correct for the color error introduced by the one -line delay.
8 -1. THE RCA DOC SYSTEM

The color dropout compensation system, as installed in the RCA TR -70,
is shown in Fig. 8 -1. The system consists of the Dropout Compensator
module (620) , the dropout processor module (619) , and a delay line
(1A3) The dropout processor module utilizes a submodule board called
.
the sample -pulse eliminator. This is installed in the ATC error detector
module (623) .
The dropout fm input to the dropout processor ( from the fm equalizer

module) is filtered and passed through an automatic gain-control stage.
From here, the dropout fm is fed to the Dropout Compensator module,
where dropout detection takes place. At the same time, video from the
post- emphasis module (529) is fed to the video switch in the Dropout
Compensator module. The video signal, after passing through the video
switch, is fed simultaneously to a 63- microsecond delay line and back to
the post-emphasis module. If no dropout is detected in the fm signal,
video continues to pass uninterrupted through the video switch and back
to the post-emphasis module. Thus, in the absence of a dropout, the sub-
sequent video output of the post- emphasis circuit is unchanged.
If the fm signal level drops below a predetermined level, the Dropout
Compensator module produces a dropout pulse. The fact that a dropout is
detected in the fm signal indicates that the counterpart segment is missing
from the post-emphasis video. The dropout pulse activates the video
switch, causing it to interrupt the incoming video and to send to the post -
emphasis module video from the preceding scan line, which is stored in
the delay line. Thus the lost segment of video is replaced from the previ-
ous line of video.
The dropout pulse is also looped throughout the dropout processor
module to the sync -channel video section in the post-emphasis module.
This pulse, called the sync- dropout pulse, interrupts the sync -channel video,
thus preventing noise caused by the dropout from triggering the sync
circuits.
If the dropout takes place during horizontal sync, the sync -dropout
pulse, after being applied to the dropout processor module, is stretched by
two thirds of a line prior to being fed to the sync -channel video circuit.
In this case, a coincidence detector triggers a generator which, in turn,
produces a 42- microsecond pulse that is added to the existing dropout
pulse. This stretched sync -dropout pulse eliminates horizontal sync plus
RI 21i
...1.-
E-E FM
From R2
FM Ref
508 FM 16 Sensitiv ty
TP3 Set
Ref Frame From
Cont Track
27
Rec/Play
617
Sim DO
TP2
9 I
Cancel From
6
PfaylRec
Panel
Sìmmi
Dropout
Insert 7
Lead Edge Gate +20V
Horiz
From
28 Gate
Int Ref On Voltage
621
32
TP1
30
31 17 Dropout Processor 619
DO Pulse DO FM
NC PIL Bus Play Bus
16181 16181
FM FM In From
19 17
Playback Amplifiers
17
Video FM To Sync FM To
FM Filter Demodulator
15241
Fig. 8-1. Functional diagram

DO FM 20 Dropout 17 Video In
Compensator From
DO Det
620 Video Filter
DOC 15 18 527
Cont Video Switch 18 Pio Video
32
27 Video
DOC Out
Video DOC
29 Function
Interlock
Switch
63 Vs
D lay Li e
Prior to
DO Sta d
10
DOC
Interlock Video Pix Video
M2 Driver Out To
Pair 31 MATC Video
624
Tape Sync Proc
622 20
T pe Vert Split
16
a
31A_ 17
14
Sync Post
Charm I
Emphasis
Tape Vert Ref Vert From
Video 529
Ref Gen 613
6
DO 28 ¡20 4 5
Vert ps Del y
2 31
Sync Sync Channel
Coin Gate And Filt r Demod "*-Tape FM I n
1A3 530 17
1 30 From FM Switch
Linelock 522
610
Sync DO Pulse
27
Sample Sample
AND
Pulse Pulse
Gen Elim Gate
ATC E ra Detector
17 623 26
L
MATC Sam le Pulse In 9H In From

From Tape Sync Proc Sync Logic
622 533
of dropout compensation system.

a half line of video. If the dropout pulse is derived at horizontal -sync

time during the 9H interval, it will remove the succeeding equalizing
pulse. The sync -dropout pulse is fed simultaneously to the sample -pulse
eliminator in the ATC error detector module, along with the 9H pulse
from the sync logic module. When coincidence between both of these
pulses exists, ATC operation is suspended.
The same stretched pulse is used to drive a simulated vertical-sync
pulse called the dropout vertical. The dropout vertical is fed to a coinci-
dence gate (in the Linelock module) that controls the generation of
vertical- timing pulses. When Pixlock is achieved, the dropout vertical
serves as a substitute pulse in the event that the tape vertical -sync pulse,
normally used in timing, is lost due to a dropout.
The delay line, 1A3, delays the sync video to allow time for the Drop-
out Compensator to sense dropouts. If this delay were not present, the
leading edge of the dropout would be able to pass without compensa-
tion, and noise would appear in the picture. The passband of the delay line
is limited to 2 megahertz, resulting in the elimination of all chroma in-
formation in the sync -channel video.
The sensitivity control is used to establish the threshold level of drop-
out detection. For test purposes, simulated dropouts, which are visible on
a monitor, can be inserted into the video in either play or E -E operation.
The Dropout Processor Module

The circuitry of the dropout processor module can be divided into three
sections, each broadly classified according to function. (See block dia-
gram, Fig. 8 -2) .
1. Filtering, automatic gain control, and amplification of the dropout

(DO) signal.
2. Simulated dropout generation whereby the functioning of the Drop-
out Compensator system as a whole can be visually evaluated on a
monitor. Included is the sensitivity set circuit. This circuit, in con-
junction with the sensitivity circuit in the Dropout Compensator
module, determines the fm level at which dropouts will be selected.
3. The dropout pulse circuit, which includes the sample-pulse elimi-
nator in the ATC error detector module. The dropout pulse circuitry
provides pulses to supplant primary timing pulses that are lost be-
cause of dropouts in the fm signal. The dropout pulses either inhibit
circuits used in generating timing pulses for the servo systems or
replace missing horizontal- and vertical -sync pulses.
In the play mode, the tape DO fm signal is passed by the input - signal
selection relay (K2) to drive a low -pass filter which eliminates high -
frequency noise. In the E -E mode, the relay selects the fm reference sig-
nal, which prior to filtering is fed through a feedback amplifier. Ampli-
fication of the fm reference signal is necessary in order to drive the filter.
The filtered fm signal is applied to the agc circuit across two field -effect
transistors, and it is coupled by an emitter follower to the first of two
feedback amplifiers.
The feedback amplifiers provide a gain of 36 dB. The field -effect tran-
sistors act as electronically variable attenuators in the path of the fm
signal. The field -effect transistors are controlled by the dc amplifier, and
their impedance is inversely proportional to the magnitude of the recti-
fied feedback signal. If the amplitude of the fm signal tends to increase,
the impedance of the field -effect transistors is lessened, and more of the
fm signal is attenuated. Conversely, a reduction in the amplitude of the
fm signal results in an increase in the impedance of the field -effect tran-
sistors, which in turn results in less attenuation. Consequently, the level
of the fm signal being fed to the Dropout Compensator module remains
constant, despite fm level variations such as those encountered between
different fm standards or tape recordings. Amplitude control of the fm
signal is such that the output of the agc circuit remains within ± 1 dB
with an input change of -!-10 dB. Since the agc circuit responds only to
very slow changes in the amplitude of the fm signal, dropouts do not
affect it.
The dropout pulse, derived when a dropout is detected in the fm signal,
is at least seven microseconds long; it may be longer, depending on the
duration of the fm dropout. On being applied to the dropout processor
module, the dropout pulse may follow either of two circuit paths, depend-
ing on when it occurs in relation to the sync or video component. If the
dropout pulse is derived between sync pulses, the dropout pulse is fed
directly through the dropout processor to the post-emphasis module. The
dropout pulse is applied to the DO clamp in the sync- channel video cir-
cuit, and, as a result, sync video is clamped at gray level during the
dropout.
If a dropout takes place during sync, the dropout pulse fed to the
dropout processor module is coincident at an AND gate with a 17- micro-
second pulse derived from the inverted leading -edge gate pulse. When
this happens, the output of the AND gate triggers the 2H/3 multivibrator.
The multivibrator in turn produces a pulse lasting two thirds of a line,
or 42 microseconds. This pulse added to the dropout pulse becomes the
stretched sync dropout pulse, which is fed to the DO clamp circuit in the
sync -channel video section of the post-emphasis module. The stretched
dropout pulse eliminates horizontal sync or, during the vertical interval,
two equalizing pulses.
Besides being fed to the post- emphasis module, the sync dropout pulse,
stretched or otherwise, is also fed to the sample -pulse eliminator in the
ATC error detector module. The sync dropout pulse is one of the two in-
puts to an AND gate; the other is the 9H pulse. The sample -pulse elimi-
nator produces a pulse that inhibits the ATC sample -pulse generator if a
dropout is present during the 9H interval. Thus, whenever the sync drop-
7:i7 TELEVISION BROADCASTING
DO FM
i
I n put FET FET
TP3
(Tape FM) Q8 Q7
FM In
(From FM Equal 523)
FM +Ref Input vVe bVr

(E
FB Amp
015, 016
-a 014
EF
-oJ!
'
(From FM
Ref 508)-
FM +
-26 V I I
-11 MHz
Ref Out Low -Pass Filter
EE
(To FM Filte 524)
Play AND Gate

Bus X6. X17, X18
52
+20 V-4-0
J_v-_
Dropout
DO Test
(Play
Control
- Inverter
042
Insert
BSMV
Q43, Q44
X19
N
Panel)
Coincidence Circuit
Ref Frame Delay MSMV X20

Input w I n verter
MV 10H
(Control - Q49
Q47, 048 Q45, 046
-
Track Red
Play 617)
Leading Delay MV MSMV X21

EF (n verter
Edge w 78 us 30 us
Q37 Q36
Gate Input Q34, Q35 Q32, Q33
(From Int
Ref 621)
i
DO Pulse MSMV
Input _ w . w
EF
041
AND Gate EF
2H/3
Q39, Q40 038
(From Q26, Q27
DO Comp
620)
Sync DO Pulse
TP1
DO Pulse
PIay1LL Bus -, AND Gate

Q28
Amp
Q29
Fig. 8 -2. Block diagram

THE TAPE - DROPOUT COMPENSATION SYSTEM "l i
AGC Control
DC Amp
013
DO FM Out
TO DO Comp 620)
EF FB Amp AGC Del
Q6 Q4, Q5 X7
R49
FB Amp EF
QQ, Q3
Ql Q12
-26 V-- DOC

Control
..e_dGate Driver (To DO
K1 hll
Q11 Comp
620)
S1
mi/ Sensitivity
Set 1
1 TP2
Sim DO
o
- Gate Driver
Q10
Phase
Split
Q9
Clamp
R45
Balance
AND Gate EF
WI Q30
9H In
(From Sync Logic 533)
Sync DO
On Voltage
ITo Post -Emp 529)
Sample Pulse Eliminator

Sub -Module In ATC Error Det 623
EF A3
Q2
Coincidence Circuit
X2
X3 Q9 -C
Al
Inverter AND Gate
Q3
BSMV
Gate
04
-ro-
Q1
DO Vert
ITo linelock 6101
of dropout processor module.

out pulse and the 9H pulse are coincident, the ATC sample pulse is elimi-
nated. This allows the ATC circuits to continue with their previous com-
pensation.
The 2H/3 multivibrator also generates another pulse from which is
derived a dropout vertical pulse. This pulse, which is fed to the Linelock
module, replaces the vertical -sync pulse normally used in vertical timing in
the event it is lost due to a dropout. When Pixlock is achieved, the Pix-
lock /Linelock bus is at ground potential. This potential and the 2H/3-
generator pulse furnish the inputs to an AND gate. During the period that
coincidence exists between the AND gate inputs, the AND gate is enabled,
and the succeeding amplifier produces the dropout vertical pulse.
The dropout vertical pulse is applied simultaneously with the tape ver-
tical signal to the tape -vertical /reference coincidence detector in the Line-
lock module. Whenever the vertical -sync pulse that is used in deriving the
tape vertical signal is eliminated due to a dropout, the dropout vertical
pulse is present as a substitute. Thus, the servo systems are prevented from
falling out of lock during a dropout in the tape vertical sync.
The simulated -dropout circuit permits the operation of the dropout -
compensation system to be checked by inserting dropouts in the fm sig-
nal. The simulated dropouts are inserted just prior to the fm output stage
that feeds the line to the Dropout Compensator module. Depending on
which signal, fm or video, is selected for observation, the dropouts will
appear as gaps in the continuous fm wave or as lines in the video. The
simulated-dropout insertion circuitry is activated by pressing the DROPOUT
INSERT switch on the front panel of the dropout processor module. To
view the dropouts in the signal on the cro or picture monitor, the COMP
OFF /MONO ONLY /NORMAL switch on the front panel of the Dropout
Compensator module must be in the Mono Only position.
The dropouts are placed in the fm signal by a diode quad (M1) which
is gated on during five alternate lines once in every frame. The quad gat-
ing signals are derived from the output of a three -input AND gate. One
of these input signals is a steady -state control signal. The second is a 10H
pulse timed to occur about four milliseconds after the beginning of each
frame. The third is a 30-microsecond pulse timed at half the horizontal
rate to occur about five microseconds after the start of a line. Conse-
quently, once every frame for a 10H period, all three AND -gate inputs will
be coincident. During this period, five 30- microsecond pulses (one every
other line) will be passed by the AND gate. These pulses are then split
in phase to become the quad gating pulses. The simulated dropout passed
each time the quad is turned on causes a reduction in the level of the fm
signal being fed to the Dropout Compensator module. This reduction in
level is sufficient to activate the dropout- detection circuitry in the latter
module. When the demodulated signal is viewed on a picture monitor,
five lines about one quarter of the way down the raster will be visible in
the display.
The dropouts are eliminated by resetting the flip -flop that produces the
steady -state input to the AND gate. This is done by tapping any of the
following switches: STOP, WIND, PLAY, STANDBY, RECORD, SETUP, CUE
ON /OFF, or AUDIO ON /OFF.
The DOC Module

The Color Dropout Compensator (DOC) normally requires no atten-
tion by the operator and functions automatically whenever power is ap-
plied. However, there are two operational controls that affect the opera-
tion of the Dropout Compensator, and their settings should be considered
by the operator. They are the SENSITIVITY control and the mode selector
switch (Fig. 8 -3) Their functions are explained in the following para-
.
graphs.
The SENSITIVITY control adjusts the sensitivity of the dropout detector
circuit. To the detector, a dropout is a reduction in the rf amplitude of the
fm carrier. As the SENSITIVITY control is turned clockwise, the sensitivity
decreases so that a greater drop in the rf amplitude is required to cause
the video to be replaced. As the control is turned counterclockwise, the
smaller a drop in the rf amplitude must be to cause the video to be re-
placed. The useful range of the SENSITIVITY control is the upper (clock-
wise) two- thirds of its range. This is 10 to 20 dB below the normal rf in-
put level of 0.4 volt. As the SENSITIVITY control approaches the maximum
counterclockwise position, the detector sensitivity becomes so great that it
causes excessive replacement of the video.
When determining the sensitivity setting, the operator must consider
the condition of the tape and the condition of the recorder. Amplitude
variations caused by the recorder, such as those resulting from slope detec-
tion or tracking variations, could be detected as dropouts if the sensitivity
is too great. The ideal sensitivity setting is one which compensates for all
dropouts, large and small, up to the point where the replacement video
causes an apparent change in the overall picture.
In general, if the rf signal level is consistent, operation at a higher
sensitivity is permissible. If the level is not consistent, it is preferable to
operate at a lower sensitivity. Under noisy signal conditions, the DOC
will, of course, supply more replacement video. If the sensitivity is too
great, the difference between the replacement video and the direct video
could be perceptible in the overall picture.
The mode selector switch has three positions that correspond to three
modes of operation: Normal (normal full compensation of NTSC color
signals or black and white) , Mono Only (monochrome only) , and No
Comp (no compensation) These functions are explained in the follow-
.
ing paragraphs.
In the Normal position, all functions of the dropout compensator are
operating. This is the position that should normally be used for 525 -line
or 625 -line NTSC color or for monochrome operation. The dropout corn-
pensator will replace both the luminance and the color subcarrier with the
stored video, in the event the reproduced rf level drops below that set by
the SENSITIVITY control. If the rf level is normal, the video is routed di-
rectly through the video switch to the post amplifier and the video output.
In the Mono Only position, the DOC operates as in the Normal position,
except that the color subcarrier is removed from the replacement video.
The direct video is not affected. That is, as long as there are no dropouts
and the DOC is supplying direct video, the chrominance signal and color
burst are unaffected. The color signal is eliminated only in the delayed
video inserted in the event of a dropout. The Mono Only position would
normally be used with color systems other than the NTSC color system.
In the No Comp position, direct video is always supplied from the
DOC. The detector circuit is locked in the no-dropout state, and the
video is routed through the direct -video side of the video switch and
amplified to unity gain in the post amplifier.
Operation of the Color Dropout Compensator is based on the concept
that lost information may be replaced by information from the preceding
scan line with little or no noticeable degradation of the picture. To do
this, the dropout compensator always has stored in its delay system the
previous scan line of video information. By continuous sampling of the rf
level of the reproduced fm signal, it may be determined if the video has
a significant loss of information, or if it is acceptable.
If the rf amplitude of the reproduced video signal is acceptable, as set
by the SENSITIVITY control, the signal is simply routed through the DOC
and back to the recorder output circuitry. At the same time, the acceptable
line of video information is sent through the delay system to be ready for
use if the following line has a significant dropout. If the rf level drops
below the level set by the SENSITIVITY control, the DOC switches from the
current video information to the previous line of video information, which
is stored in the delay system.
The direct uncompensated video is fed into one side of the video switch
circuit, and the delayed video is fed into the other side. The detector con-
trols which video is the output from the DOC. The video output is con-
tinuously fed into the delay system and stored for one line.
Fig. 8 -3 illustrates the layout of the DOC circuit boards and the signal
flow between them. All incoming signals, except the VI bus which con-
trols K1, come into the detector and video switch board. These input
signals are the uncompensated video, the fm input, and the DOC- enable
signal from the play bus. The detector monitors the fm for dropouts and
controls the video switch as previously described. For the detector to op-
erate, a ground enable signal must be provided on pin 15 by the play bus.
Also, as previously described, the detector provides a sync clamp pulse on
pin 29 when a dropout is detected. The video switch, controlled by the
detector, selects which video, the direct video from pin 18 or the delayed
video, will be the output.
a
F
ó
>
co
á
o
Ti
o' É
L
ó
> o
>
G
E
o >
o
á
or
U
G E
o
m
U
E r
c` á
Courtesy RCA
Fig. 8 -3. Block diagram of color dropout compensator.
The output from the video switch goes to two places, the post amplifier
board and the luminance delay modulator board. The post amplifier am-
plifies the video output signal to provide an overall gain of unity through
the DOC. The video signal to the luminance delay modulator board goes
to relay K1. When the system is operating under the 525 -line NTSC sys-
tem, K1 is de- energized and the video signal is passed straight through.
When the system is operating under the 625 -line NTSC system, K1 is
energized, and the video signal passes through a 0.5- microsecond delay
line. From the relay and delay line, the signal goes to two places, the lumi-
nance delay modulator and the chroma delay board.
On the luminance modulator board, the video information is modu-
lated onto a 10 -MHz carrier. From the luminance delay modulator board,
the modulated signal goes to the luminance delay demodulator board,
where it is delayed one scan line of time and then demodulated. The de-
layed luminance then goes to the mixer circuit on the chroma delay board
to be mixed with the delayed color information.
On the chroma delay board, the video signal from K1 on the luminance
delay modulator board is filtered so that only the chrominance informa-
tion remains. The resulting signal is phase inverted and then delayed one
RF In
(From Demod Input Gate)
r
Dropout Sensor
R58 Dropout
Detector
R60
Level T1
Emitter
Limiter
Q12
Fol lower
Q15
lirDetecto 018
Low Pass
Filter
L1,L3,C20,C3
Emitter R42
Lim'ter Emitter
Follower Amplifier
Q14 Fol lower
Q16
013 Q19
P43
o Play
Emitter
0
Follower
Q10
oy
X6 Record
Tunnel
Diode
K2
Switching Signal Out Amplifier
10
To Demodulator Q11
From Cal -Pulse Gen
Fig. 8 -4. Block diagram of dropout sensor section.

scan line of time. The chrominance information is then mixed with the
luminance information, and the complete video signal, delayed one line,
is fed from the chroma delay board to the delayed side of the video switch
on the detector and video switch board.
The reason the luminance and color are delayed separately is that the
color signal of each succeeding scan line is shifted in phase by 180 °.
Therefore, the delayed video must have its color subcarrier not only de-
layed one line, but also phase inverted, as previously explained.
8 -2. THE AMPEX DROPOUT COMPENSATOR

Module 6 of the Ampex VR -2000 machine includes two principal sec-
tions, the outputs of which are relay -selected as a function of the system
operating mode (i.e., play or record /E -E) The two sections share a two -
.
stage output amplifier; see Figs. 8 -4 and 8 -5. (One of the sections is the
calibration -pulse generator, which is not discussed here.)
There is a relay- selected rf signal at pin 6 of the demodulator input
gate, and this signal is applied to an input of the dropout sensor. This rf
signal (approximately I/2 volt peak to peak) enters the dropout sensor
(module 6) at pin 22 and is routed through dropout- detect level control
R58 to the base of limiter Q12. The output of Q12 is coupled by emitter
follower Q13 to the base of second -stage limiter Q14. The amount of limit-
ing is a function of the signal level applied to Q12, which in turn is a
function of the setting of potentiometer R58.
The output of Q14 is coupled by emitter follower Q15 to full -wave
rectifier (or detector) Q17 -Q18 through transformer T1, which provides
signals of opposite phase at the bases of Q17 and Q18. Because both are
pnp transistors, when one turns on, the other turns off, resulting in full -
wave rectification at their common collectors. As long as the level of the
incoming rf equals or exceeds the limiting level, the output of Q17 and
Q18 is a straight line. If the level of the incoming rf falls below the
limiting level, both transistors turn off, and their collectors swing toward
ground (zero volts) .
The limited and full- wave -rectified signal at the output of Q17 -Q18 is
routed through a low -pass filter to the base of emitter follower Q19. The
emitter follower remains cut off as long as the positive voltage of the sig-
nal from the collectors of Q17 and Q18 remains at the limiting level.
When, however, this voltage falls below the limiting level, Q19 turns on,
and its emitter voltage swings negative. The result is a negative pulse the
duration of which coincides with the duration of the dropout period
(Fig. 8 -6)
.
During the play mode, the output of Q19 is routed through de- energized
relay K2 and coupled by emitter follower Q10 to the base of switch Q11.
The base circuit of Q10 includes tunnel diode X6, which acts to greatly
accelerate the rise times of the negative dropout pulses applied to the
base of Q10. Switch Q11 is normally off, but it is turned on by each

negative dropout pulse. Its output therefore remains at -12 volts until a
dropout pulse occurs, when the output swings to zero volts. The leading
and trailing edges, respectively, of each dropout pulse turn Q11 on and
off. (Refer to Fig. 8-6.)
The collector output (positive -going pulses) of Q11 is routed to pin
10 and is the switching signal received at pin 21 of the demodulator in-
put gate. The output of Q11 is also applied to the base of hysteresis gen-
erator Q16, which functions as a switch. The rf present in the circuit of
limiter Q14 includes a small amount of amplitude modulation, which
could make the indication of a marginal dropout somewhat ambiguous,
rather than positive. As noted earlier, Q11 turns on during a dropout
period, and its collector swings (positive) to zero volts. This positive volt-
age swing is applied to the base of normally on switch Q16 and turns
it off.
+12V +12V
18
18 Limiter
R39 100 R46 620
R56
C34 0.1 R55 1000 C33 0.1

C25 0.1
Q12 013 C32 Q15
20 20 _ 2N708 20 2N708
2N708
C31 0.01
R58
Dropout 200 0.1
Det Level
10%
X7 X8 L4 - 330 X9 X10 L5 - 330
1W R59 4700
013-972 013-972 pH 013 -972 013 -972 PH
R40 5600
20 - 20= 20= 20= 20 = 20= 20
19
75 p -12V C17
100pf
C24 500 V 500V 39 pF R45 1200
R42 150
838 75
R41 22
C21500V
20 - R43 P0- +12V
19
-12V
18
240 C28
C23 837 1500 C27 944 2200 C18 0.1 0.01
I0.1 I.1 R47 24k
20 -_ 19 -_ 19
20 Q16
-12V -12V
202048
Hysteresis
Generator
20 =
19 16 18
R50 R49
47 6800
41
C16 47 C6 15
I20V T 20V
20 20 = 9 =
22
RF In Gnd G d -12VDC Not U ed +12 VDC
Fig. 8 -5. Schematic diagram

When the rf level equals or exceeds the limiting level, Q16 is satu-
rated, and the effective emitter resistor for limiter Q14 is R42 (150
ohms). Thus, under this condition the gain of Q14 is approximately 10
(i.e., R56, 1500 ohms, divided by R42, 150 ohms). When the rf level
falls below the limiting level (i.e., the dropout condition) , Q11 turns on
and biases Q16 off. Under this condition, the effective emitter resistor for
Q14 includes R42 plus R43 (240 ohms), and the gain is 1500/390, or
approximately 3.8.
When the gain of Q14 falls from 10 to 3.8, the signal level at the base
of Q15 falls, removing any possible ambiguity in the indication of a
dropout that is marginal. Thus the presence of amplitude modulation
cannot cause a marginal dropout to appear to be a cyclic alternation be-
tween a normal and a dropout condition. When any dropout occurs (e.g.,
rf level down 20 dB from normal), Q11 will remain on, Q16 will remain
off, and the gain of Q14 will remain at approximately 3.8 until the in-
TI
Dropout Detector
018
202048
+12V
R48f 47
18
1
C29
3pF500V
L3
39 pH
C19
12vF500V
LI
27 pH
E
Pulse Stretcher
C21
20 =
OI
1OV
R33
12 V
18
5600
R34
Q19
2N2048
C30_68p1 C20 R54 1300

I500V 39 pr R51 3900
ROO
43
20 = 20
- 500V
20 =
19
-12V
R57 C22 R35 100

Switching Pulse farmer 11.1
R53 TPI +12V 20 - -12V

16
18
fill R32 75
Q10
2N2048
2N1308
R31 1000
X6
K2
114371 :
R26 5600
936 820
16 - 9
-I2V
47
R52}47 C14
20V
19
-1 V
1171151141131 EI El } 11 1E1
Nd Used Gnd Cod Switching From EE Tape
Signal Out Cal -Pulse Control
Section of In
Board 6
of dropout sensor.
coming rf level is down only 11 dB from normal. At this point Q11 will
turn off, Q16 will turn on, and the gain of Q14 will rise to the normal
level of 10.
Dropout
Condition
Output of 015
RF In (Pin 22( Bases of 017 -018)
Base of 019
Filtered Output
Limiting Level (0 V)
of Q17 -Q18
019 Off
Emitter of 019
Q19 On
011 On (0V)
Collector of 011
011 Off (-12 V)
Fig. 8 -6. Waveforms in dropout sensor.
EXERCISES
Q8 -1. Define the term "dropout."

Q8 -2. What is the basic reference for a tape- dropout compensation system?
Q8 -3. If an original (first -generation) recording contains dropouts, will a dub
of this recording contain the same dropouts?
Q8 -4. When a dropout occurs in a given line of picture information, what
happens?
Q8 -5. What is the basic additional function for color DOC as compared to
monochrome DOC?
CHAPTER 9
Advanced Color -Error

Correction Systems
Chapter 6 traced the need for special compensation systems for mono-
chrome geometric error, and the additional types of correction necessary
for the color information. These corrections act primarily on a head-to-
head basis, or on the entire group of 16 to 17 lines occurring within one
head pass across the tape.
There are additional color errors which are not "cleaned up" by the
previously discussed systems. These are:
1. Hue (phase) shifts within each band of 16 to 17 lines of a head
scan.
2. Saturation (amplitude) variations within each band of 16 to 17
lines of a head scan.
The advanced color -error compensation systems required to correct these
errors are the Velocity Compensator, used in conjunction with the Colortec
and Amtec (Ampex) , and the Cavec (Color Amplitude and Velocity
Error Compensation), used in conjunction with the monochrome and
Color ATC systems (RCA) .
9 -1. THE NEED FOR ADVANCED COLOR -ERROR

CORRECTION SYSTEMS
The two major types of errors to be compensated, hue shifts and satura-
tion variations, are produced by mechanical and operational deficiencies
that cannot be eliminated by normal machine adjustment. These defi-
ciencies are inherent in this type of equipment and have existed since the
introduction of quadruplex tape. It is only during color operation that
their effects on tape quality become apparent in the form of color band-
ing. Banding usually occurs during the passage of one head over the tape
surface.
253
Hue banding is commonly referred to as velocity error. It results from

differences in head scanning velocity between the recording machine and
the playback machine. This can be caused by many small factors, such as
different tape tension between record and replay machines, different posi-
tions of some of the fixtures on the tape transport, or incorrect guide
height adjustment.
As one specific example of saturation banding, review the text asso-
ciated with Fig. 2 -13, Chapter 2. Note that head -to -tape pressure varies
between top, center, and bottom of the tape when perfect concentricity
does not exist. (Recall that "perfect" concentricity is practically impossible
to achieve on a tape -interchange basis.) This change in pressure causes a
change in the high -frequency response of the head, due largely to the fact
that in spite of the penetration of the pole tip into the tape, a minute air
film exists between the head and tape. Pressure changes across the arc of
the head path cause a modulation of this air film, which affects the high -
frequency response.
For under -penetration (Fig. 2 -13B), pressure is lowest at the center of
the head pass. At top and bottom, the head is closest to the tape and the
air film is smallest. Under these conditions, the carrier frequency itself is
reproduced well, but there is a reduction of chrominance (sideband) sig-
nal level relative to that at the center of the tape. At the tape center, the
lower pressure causes the air gap to increase, lessening the fm high -
frequency response and increasing the chrominance signal level relative
to the level at the top and bottom of the head-scan arc across the tape.
The net result of this defect is color banding within each band. Not
only does the chroma amplitude change across the complete head scan
of 16 to 17 lines, but differential gain and differential phase occur also.
Therefore, chroma, or saturation, banding is the result of differences in
head -to -head frequency response, tape surface variations, and tape -to -head
contact differences. The effect is a change of chroma amplitude between
the heads, as well as a saturation change within any one head pass. In
simple terms, the chroma amplitude varies from one line to the next. This
requires compensation on a line -by -line basis. It is possible to correct
manually only for individual head -scan (16-line) intervals; thus the cor-
rection is only an averaging for line -to -line errors. Electronic compensa-
tion, however, is capable of correcting the actual error, line by line.
Some of the errors present in tv tape recording can be compensated for
by manual adjustments in the playback mode. This is true for both
chrominance- amplitude errors and velocity errors, but it is difficult to
remove both effects simultaneously, and in removing one effect the oper-
ator may cause another. An example of this is evident if an interchange
recording with objectionable saturation banding is played back. Mechani-
cal adjustments such as guide height may be made, and electrical adjust-
ments on individual heads may be made manually by turning the play-
back equalizers in the individual channels. The saturation banding might
ADVANCED COLOR -ERROR CORRECTION SYSTEMS 255
be compensated for perfectly, but in doing so, velocity errors would be

introduced. Velocity errors are seen as a continuous change in color phase
across the tv line or down the head band. The result is that the picture
probably looks worse than it did with the original saturation-banding
error. What all this adds up to is that no matter what is done with the
original saturation -banding error or how many adjustments are made, the
end result of playing back a color interchange recording that had errors
in some form is color banding seen by the viewer.
9 -2. BASIC FUNCTION

The basic function of the Cavec unit (Fig. 9 -1) may be described by
dividing it into three basic parts:
1. Integrated- circuit digital memory system from which the line -by -line
correction is obtained
2. The velocity or phase error detector (VEC)
3. Chrominance amplitude corrector (CAC)
The line -by -line correction is made possible by the basic assumption
that errors found in one revolution of the headwheel will also be present
in the next revolution of the headwheel. During the head revolution, each
of the four heads scans 16 lines; thus, to store information for one com-
plete revolution, 64 memory units are required. Separate memory units are
required for the velocity and chroma-amplitude corrections. Thus, the
memory units in any one Cavec total 128. The memory units are really
capacitors which are charged in approximately 100 milliseconds, or 24
revolutions of the headwheel. It should be recognized that instantaneous
changes in either hue or saturation will not be corrected to some absolute
value but will be corrected to the average value of the previous 24 revo-
lutions of the headwheel.
The CAC corrects for frequency -response variations in the video sig-
nal, which produce saturation errors in the color picture. The VEC cor-
rects for inaccuracies in head-to -tape velocity, which cause undesired hue
changes.
The circuits of the Cavec are similar to those used in computers. Some
are of the analog- computer type, and others are of the digital- computer
type. As shown in Fig. 9 -1, the analog circuits are divided into two
groups, one for CAC and the other for VEC. The digital circuits are a
single group shared by both VEC and CAC. The output signals which
correct the errors are produced by the analog circuits. The digital circuits
provide the analog circuits with timing and control pulses required for
line -by -line correction.
The input to the CAC (Fig. 9 -2) consists of burst separated from the
demodulated video. In the analog circuits, the burst amplitude is com-
pared to a reference level, and the resulting error voltage is stored in a
o
ii
E
G
o
E
ó
dlñ°5
CJV QCJO
iiit-
E
o
- x
Fig. 9 -1. Simplified block diagram of Cavec unit.

capacitance memory. The memory contains 64 capacitors, one for each

line in the four head scans; thus, 64 errors are stored. At the beginning
of each line, the stored error voltage corresponding to the line is read out
and converted into two push -pull voltages which constitute the CAC in-
put signals. These signals are fed out of the Cavec to an electrically con-
trolled attenuator in the fm equalizer module. The attenuator automati-
cally adjusts the fm equalization in a direction that tends to reduce changes
in the burst amplitude. Correction occurs before the portion of the fm
signal containing the burst arrives at the fm equalizer, thereby providing
true closed -loop operation. Thus, the CAC resembles an automatic gain-
control system.
Tape FM In-. FM
Equalizer
Module
FM
Demodulator Tape Video
±V Equalizer
Control Voltages
Filtered
Cavec Burst
Unit Filter
(CAC)
Burst Separator
Fig. 9 -2. Simplified block diagram of CAC.
The VEC portion improves the monochrome ATC system of the ma-
chine by causing the MATC variable delay line to correct timing errors
continuously instead of in steps. In machines without Cavec, the MATC
error signal is applied to the variable delay line at the beginning of each
line and remains constant throughout the line. The error waveform there-
fore appears as a series of steps, as shown in Fig. 9 -3. This type of signal
restores video timing to normal at the beginning of the line, but it allows
errors to accumulate during the line. As a result, the chroma shifts be-
tween correction steps, causing hue banding or, in the case of PAL opera-
tion, desaturation.
To correct this deficiency, the VEC adds ramps ( Fig. 9 -3) to the MATC
error signal. The resulting waveform closely approximates the true, con-
tinuously variable velocity -error signal. To produce ramps of the correct
slope, the VEC analog circuits measure the differences between the suc-
cessive step voltages and store the difference voltages in a capacitor mem-
ory similar to the one used by the CAC. The memory contains 6.4 capaci-
tors, one for each of the 16 lines in each of the four head scanning periods.
(Since some head scan periods contain seventeen lines, the error stored in
the sixteenth capacitor is used for both the sixteenth and seventeenth
lines.) The time constant of the capacitor charging circuit causes the
error voltage to stabilize after 24 headwheel revolutions. Each stored volt-
age, therefore, is an average for 24 revolutions. At the beginning of each
line, the corresponding stored voltage is applied to a circuit which gener-

ates a ramp having a final height equal to the error. Each ramp is added
to the corresponding MATC step, and the combined signal is fed out of
the Cavec to the circuits which control the MATC delay line.
Although the RCA Cavec module, when mounted in the tape machine,
looks like the other modules, its internal construction is entirely differ-
ent. Instead of one large circuit board, the Cavec unit contains a number
of small printed circuit boards plugged, at right angles, into connectors
on a wiring backplane which extends for the entire length of the assem-
bly. The backplane has connectors, each containing 45 pins, for a maxi-
mum of 32 boards.
Actual
A
Timing Error
MATC Error
VEC Ramp
VEC Output
Expanded View
(Sum of
D
Waveforms
B and Cl
_17-Line 16 -Line
Head Scan Head Scan
Fig. 9 -3. VEC waveforms when guide pressure is misad'usted.
The digital system occupies the first 12 board positions. This system
actually contains 11 boards, but only one of the first two positions is used,
position 2 in the TR -70 and position 1 in all other machines. Only five
different types of digital boards are used. The principal components on
these boards are integrated- circuit packages of the dual in -line type. The
packages contain logic circuits, such as gates and inverters, or timing cir-
cuits, such as one -shots or flip-flops. All logic elements are of the diode -
transistor logic (DTL) family.
Positions 15 through 18 are occupied by four identical boards which
constitute the VEC memory. These boards contain storage capacitors,
transformers, transistors, and semiconductor switches. A total of 64 stor-
age capacitors is provided, 16 on each board. Each capacitor stores infor-
mation pertaining to a particular line in a particular head scanning inter-
val (for example, line 1 of the head -1 interval).
Positions 13, 14, and 19 contain the VEC analog boards. The compo-
nents on these boards consist mostly of semiconductor switches and inte-
grated- circuit operational amplifiers. The operational amplifiers are basi-
cally high -gain dc amplifiers using large amounts of negative feedback.

By connection of various external circuits in the feedback loop, these am-
plifiers are made to perform a wide variety of functions, such as addition,
integration, and inversion of signals.
Positions 23, 24, 29, 30, and 31 contain the CAC analog boards, and
positions 25 through 28 contain the CAC memory boards. These boards
contain components similar to those on the VEC analog and memory
boards.
A regulator providing outputs of +4.5 volts and +12 volts is mounted
on the back end plate of the Cavec unit.
The front panel of the Cavec module contains on /off switches for the
VEC and CAC systems, a potentiometer for manual control of the CAC
burst ratio, and four test points. Internal screwdriver-adjustment poten-
tiometers on the printed- circuit boards can be adjusted without unplug-
ging the modules from the backplane.
To facilitate testing, a connector accessory, a board extender, and a
board extractor are provided with the Cavec module. The connector ac-
cessory, in conjunction with the standard module extender provided with
the machine, permits extending the Cavec module in its normal vertical
position so that the backplane is accessible from the left side while most
of the potentiometers on the printed- circuit boards are accessible from
the right side. The board extractor and extender permit testing compo-
nents on the individual circuit boards.
9 -3. THE VELOCITY -ERROR CORRECTOR (VEC)
NOTE: It is imperative that the reader have a background in logic cir-

cuitry at least equivalent to that contained in Chapter 3 of Television
Broadcasting: Systems Maintenance, by Harold E. Ennes (Indianapolis:
Howard W. Sams & Co., Inc., 1972).
The velocity -error corrector (VEC) contains two main groups of cir-
cuits, as shown in Fig. 9 -4. The first group, called the analog system (Fig.
9 -4A) , consists mostly of operational amplifiers and semiconductor
switches. The second group (Fig. 9 -4B) , called the digital system, contains
pulse generators, counters, gates, and other forms of digital logic circuits.
The functions of the analog system are to generate a series of ramp sig-
nals (straight -line segments) , one for each horizontal line in each of the
four video -head scanning periods, and to add the ramps to the MATC
error signals. The MATC error signal consists of a series of steps which
change level at the beginning of each line. The amplitude of each step
constitutes a measurement of the velocity error existing at the beginning
of the corresponding line. The true velocity error, however, is a smoothly
changing signal, rather than a series of steps. Addition of the ramps
smooths out the steps to provide a better approximation to this signal.
E
U
L J i
2 C
1- N
t. 2C
11
an
E E
E Q
N
Fig. 9 -4. Simplified block diagrams of VEC.

The function of the digital system is to generate a number of timing

pulses from the tape horizontal sync, 2 X 1 switcher, and 4 X 2 switcher
signals of the tape machine. These pulses control semiconductor switches in
the analog system which perform the various functions required for gen-
erating the ramps.
VEC Analog System
The peak-to -peak amplitude of the ramp which must be generated to
provide a smooth connection between steps depends on the difference
between the velocity errors existing at the beginning and end of the line.
The total velocity error must include the CATC (Color Automatic Timing
Corrector) signal as well as the MATC signal. Consequently, the informa-
tion required for generating the ramps is obtained by processing both the
MATC and CATC error signals as shown in the block diagram Fig. 9 -4.
The two error signals for each line of a head scan are combined in a sum-
ming amplifier. This circuit amplifies each signal by a different amount, to
obtain the required balance, and adds the results. For purposes of illustra-
tion, the combined MATC and CATC signals for lines 10 through 13 are
shown in the top line of the timing diagram, Fig. 9 -5. Actually there are
either 16 or 17 lines in a head scan on 525 -line standards (15 or 16 lines
on 625 -line standards).
The line -to -line difference voltages required for determining the ramp
heights are obtained by clamping the combined MATC and CATC error
signal to ground for a short time during each line. As shown in waveform
C of Fig. 9 -5, the clamping action places the level of each step at ground
during the portion of the line remaining after clamping has taken place.
The peak -to -peak amplitudes of the clamped waveform constitute the
desired error differences.
As a typical example of clamp operation, consider the error differences
for lines 10 and 11. The actual amplitude of the combined error signal for
line 10 is designated on the timing diagram as V10, and the amplitude for
line 11 is designated as V11. Throughout line 10, voltage V10 is applied
from the summing amplifier to the input side of the clamp capacitor. When
clamping occurs (during the clamp pulse, waveform B in Fig. 9 -5), the
clamp switch grounds the output side of the capacitor, allowing it to charge
to V10. When the clamp switch opens, the output side of the capacitor
remains at ground potential for the remainder of line 10, since no dis-
charge path exists. At the beginning of line 11, the voltage on the input
side of the clamp capacitor jumps from V10 to V11, and, since the voltage
across a capacitor cannot change instantaneously, the voltage on the output
side must change by the same amount as the input side. (In effect, the
change in voltage passes directly through the capacitor.) The voltage on
the output side of the capacitor, therefore, is equal to the difference be-
tween voltages V11 and V10. This difference is denoted by E10 on the
waveforms in Fig. 9 -5.
The clamped error signals are amplified and then fed to a semiconductor
switch called the write switch. Once per line, for a short interval before the
clamping operation, a pulse from the digital system closes the write switch,
thereby applying the signals to a bus (input /output, or i /o, bus) which
goes to a bank of memory capacitors. As explained later, storage of the
error signal differences in the memory bank provides information which
Sum of
MATC
And Line 10 Line 11 Line 12 Line 13
I
I
CATO
Error
Signals E10-V11-V10
i
O 1
V10 E12
V13
VII
I
1
IV12 (
CD
Clamp
Pulse
0
Ell
O
1
E10
Clamp
Output
I_ (((%%%rrrJJJ
E12
C9 Write In CIO Write In C11 Write In C12 Write In

rte/
O
Write
Pulse CIO Read Out C11 Read Out C12 Read 'Out
Read
Pulse
0
O
Ramp
E10 E11
E12
O
VEC
Output
(Ramp
& MATC)
Solid
Line
Fig. 9 -5. Simplified VEC timing diagram.

allows prediction of the slope of each ramp at the beginning of the cor-
responding line.
The memory bank contains one capacitor for each possible line-to -line
difference in a complete rotation of the headwheel. Since a maximum of
17 lines can occur during the scanning period of each of the four video
heads, resulting in a maximum of 16 error difference voltages per head
scan, 64 memory capacitors are provided in the bank.
While the write switch is closed, additional semiconductor switches in

the memory circuits select (or address) a particular one of the 64 capaci-
tors and transfer the difference voltage from the i/o bus to that capacitor.
In the example previously given, error difference E10 is applied to
capacitor C10 during line 11. Similarly, Ell and E12 are applied to Cil
and C12 during lines 12 and 13, respectively.
Theoretically, the velocity errors are repetitive from one headwheel
revolution to another so that, for example, the error for line 5 of the
scanning period of head 4 in one revolution should be the same as in the
preceding and following revolutions. Actually, however, small random
fluctuations occur. These fluctuations are smoothed out by making the
time constant of the memory- capacitor charging circuit about equal to the
time for 24 head passes. Thus, after a period of one time constant, the
voltage on each capacitor reaches an average value which is more repre-
sentative of the true error difference than the individual values.
To extract, or read out, the required information from the memory bank,
gates in the memory- address logic circuits address a particular memory
capacitor and connect it to the i/o bus for a period near the beginning of
each line. The particular capacitor addressed during readout is not, how-
ever, the one which will be written into during that line, but the one which
will be written into during the next line. Thus, capacitor C10 is read out
during line 10, although it is not written into until line 11. This action is
required because generation of a ramp which will connect successive
MATC steps requires advance knowledge of the slope necessary for a
ramp between those steps.
For example, in order to obtain a smooth connection of the MATC steps
for lines 10 and 11, the ramp must start at zero volts and end at -E10
volts (Fig. 9 -5) Since the slope of this ramp is proportional to the height,
.
or -E10, the value of E10 must be predicted at the start of the ramp, from
the information available at that time. This information consists of the
voltage previously stored in capacitor C10, since this capacitor will not be
written into again until the next line.
During the read -out period, the voltage from the addressed memory
capacitor is applied through the i/o bus to a driver amplifier (Fig. 9 -4).
The driver output is connected to a semiconductor switch called the sample
switch. For a portion of the readout time, the sample switch closes and
transfers the driver output voltage to a capacitor which serves as a short -
term, or temporary, memory. The sample switch and capacitor are referred
to as the sample -and-hold circuit, because the capacitor retains the sample
voltage after the sample switch opens.
The voltage from the sample-and -hold circuit is applied to an integrator
circuit, which charges a capacitor at a constant rate proportional to the
sample voltage, thus generating the required ramp. The MATC signal is
added to the ramp in a summing amplifier to provide the VEC output
signal, which is fed to the MATC circuits of the recorder.
VEC Digital System

The pulses provided by the digital system may be divided into two
types, timing pulses and memory -control pulses. The timing pulses control
the analog- circuit semiconductor switches that perform the basic functions
required for generating the ramps, such as clamping, reading, writing, and
sampling. The memory -control pulses operate the switches that address the
capacitors in the memory bank during the read and write periods.
The timing pulses are derived from the tape- horizontal, 2 X 1 switcher,
or 4 X 2 switcher signals of the machine. Table 9 -1 lists some of these
pulses and their functions.
Table 9 -1. Typical Timing Pulses From Digital System

Pulse Function
Read Establishes time in horizontal line when memory capacitors are
read out.
Write Establishes time in horizontal line when memory capacitors are
written into, or updated.
Sample Determines time within read interval when memory -capacitor
voltage on i/o bus is transferred to sample- and -hold circuit.
Ramp Reset Resets ramp generator to allow generation of new ramp.
Clamp Determines time when combined MATC and CATC input signals
are clamped to ground, to derive error difference signals.
Clock Provides horizontal -rate timing reference.
The memory -control pulses consist of two types, called X and Y signals.
For the VEC system, there are sixteen X signals (VX1 through VX16) and
four Y signals (VY 1 through VY4) One X signal is provided for each
.
horizontal line in a head scanning period, and one Y signal is provided

for each of the four head scanning periods. During the read and write
periods of a given line, the corresponding X signal causes one side of four
capacitors to be connected to the i/o bus, but the Y signal grounds the
opposite side of only one of these capacitors, thereby completing its charge
path. Thus, selection of each of the 64 capacitors in the memory is uniquely
determined by the combination of one X signal with one Y signal.
Generation of Timing Signals

Since the various timing pulses are all referenced to the tape -horizontal
signal, the basic method of generating these pulses is by connecting two
one -shot circuits in cascade. The first determines the delay of the output
pulse with respect to tape horizontal, and the second determines the width
of the output pulse. For example, assume that it is desired to generate a
pulse 3 microseconds wide, delayed by 15 microseconds from the leading
edge of tape horizontal. To accomplish this, the tape -horizontal pulse is fed
to a one -shot having a period of 15 microseconds, and the output of this

circuit is fed to another one -shot having a period of 3 microseconds
(Fig. 9 -6) .
The 1 output of a one -shot is normally at the lower of its two possible
levels, jumps to the higher level when the circuit is triggered, and returns
to the lower level at the end of the pulse period. Each of the one -shots in
Fig. 9 -6 can be triggered only by a negative -going transition. Consequently,
-i 15 us
OS1
Ji
1 0S2 1
-1.o ---0 To Other One -Shots
r
Tape Hor iz Input
If Required
15 3
o
v
O
(A) Block diagram.

Tape Honiz
15µs
OS1
"1" Output
0S2
"1" Output
(B) Waveforms.
Fig. 9 -6. Use of one -shots in cascade.
at the negative -going (leading) edge of each horizontal pulse, the first
one -shot is triggered, and its 1 output produces a positive -going pulse.
At the end of the 15- microsecond pulse period of this circuit, the 1 output
falls to the low level, causing the second one -shot to be triggered. The 1
output of this circuit then produces a positive -going pulse lasting for three
microseconds, as desired.
The preceding method is extended, in the actual equipment, by connect-
ing more than two one-shots in cascade. For example, the block diagram of
Fig. 9 -7 and the associated timing diagram of Fig. 9 -8 show in simplified
form how the clock, clamp, read X, write, and write X pulses are generated.
The tape- horizontal pulse is fed only to the first circuit in the chain,
which is called the clock one -shot. When triggered by the negative -going
edge of tape horizontal, this circuit produces a positive -going 0.5- micro-
second pulse at its 1 output and a negative -going pulse of the same width
at its 0 output. The purpose of this circuit is to produce narrower pulses
than the tape -horizontal pulses; the narrower pulses serve as more con-
venient references for the other one -shot circuits.
Both the 1 and 0 outputs of the clock circuit are used as reference. The
1 output produces a delay equal to the pulse width, or 0.5 microsecond,
because its trailing edge is negative going. The 0 output, however, pro-
duces no delay because its leading edge is negative going.
Tape Horiz -.-C T

OS
Clock
OS
-ycT
OS
Read
X
- Clock
Read X
0.5 4.3
OS
Ramp
I- Ramp Reset
rCT
Reset
-
1.9
Write
OS OS 1 Clamp
--C T Write -C T Clamp
t
OS 35 10
Write
-SCT Delay
8.5 os OS
1- +WriteX
Write X Write
Delay X
0
18 15
Fig. 9 -7. Partial block diagram of VEC timing -generator circuits.
Among other circuits (not shown in Fig. 9 -7) , the positive -going clock
pulse is applied to the read one -shot. This circuit produces a 4.3- micro-
second pulse, which is used in the read portion of the memory cycle.
The 0 output of the clock one-shot is fed to the write -delay one-shot,
which produces opposing pulses 8.5 microseconds wide at its two outputs.
As in the clock circuit, both outputs of the write -delay one -shot are used
as timing references. The 1 output produces a delay of 8.5 microseconds,
but the 0 output produces no delay.
The 1 output of the write -delay one -shot triggers the write one-shot,
which produces a positive pulse 35 microseconds wide at its 1 output. This
pulse operates the write switch of the analog system. The trailing edge of
the write pulse triggers the clamp one -shot, which has a pulse width of 10
microseconds. The 1 output of the clamp one -shot operates the clamp
switch of the analog system. The clamp switch, therefore, closes as soon
as the write switch opens, and it remains closed for 10 microseconds.
The 0 output of the write -delay one -shot is used as a timing reference
for the write-X delay one -shot, which has a period of 18 microseconds.
The 1 output of the write -X delay circuit triggers the write -X one -shot,
Time (Microseconds)
0 10 20 30 40 50 60
Tape Horiz
Clock 0.5us
Read X -4.30s _1-

Write Delay 8.5ís
Write 35 us
Clamp 10us
-t
Write X Delay 18us
Write X I- 15us -.
Fig. 9 -8. Simplified timing diagram of horizontal -rate timing pulses.
which produces a 15-microsecond pulse. This pulse is used for control of

the memory switching matrix. Note that the write -X pulse falls in the
middle of the write pulse (Fig. 9 -8) .
Generation of Memory -Control Pulses
The memory -control circuits of the digital system (see block diagram,
Fig. 9 -9) are used to produce two basic types of control signals called the
X and Y signals. These signals control the semiconductor switches that
address the memory capacitors during the read and write periods. There
are 16 different X signals and four Y signals, so that 64 pairs are possible,
J T
Le,
_ AIL
JO
Á Ñ
Ñ
>
cn
N
Ir
>
O.
LT-
Fig. 9 -9. Simplified diagram of VEC memory -control circuits.

each consisting of one X and one Y signal. Each pair corresponds to one
of the 64 capacitors in the memory bank. A particular capacitor is addressed
when its corresponding X and Y signals occur simultaneously. The X signal
closes a switch that connects one side of the capacitor to the i/o bus, and
the Y signal closes another switch that connects the other side of the ca-
pacitor to ground.
Information for generating the X signals is derived from a line counter
and a decoder. The counter provides information as to which line is being
scanned in any of the four head scanning intervals. The X decoder combines
this line -count information with the read and write pulses to determine
which capacitors should be connected to the i/o bus during the read and
write intervals. (As previously mentioned, in any one line, different capaci-
tors are addressed during the read and write periods. For example, in line
11, C11 is addressed during the read period, but C10 is addressed during
the write period.)
One Y- signal interval is provided for each of the four head scanning
intervals. Information for generating the Y signals is obtained by using a
decoder to extract from the 4 x 2 and 2 x 1 switcher outputs of the re-
corder information that indicates which of the four heads is scanning the
tape. This information is combined with the read and write pulses to
form the four Y- interval signals.
To permit addressing the memory capacitors in the desired sequence, a
switching matrix is used as shown (simplified) in Fig. 9 -10. There are 16
control buses, VX1 through VX16, from the X decoder, and four control
lines, VY1 through VY4, from the Y decoder. The prefix "V" is used to
distinguish the control lines of the VEC memory from the corresponding
lines of the CAC memory. The bar over the VY symbols indicates that the
switches controlled by these buses are closed when the signal voltages are
at the lower of the two possible levels.
An individual normally open switch is connected between one side of
each of the 64 capacitors and the i/o bus. Each VX bus simultaneously
controls four of these switches. Each VY bus controls only one switch, but
each of these switches is connected between ground and 16 capacitors, one
from each of the 16 VX groups.
Normally the VX buses are low and the VY buses are high, so that all
switches in the memory are open. When a given capacitor is addressed,
however, the corresponding VX bus goes high, thereby closing four
switches and connecting four capacitors to the i/o bus. Simultaneously,
the corresponding VY bus goes low, thereby closing the Y switch corre-
sponding to the head scanning interval, which grounds 16 capacitors.
Only one of the 16 capacitors, however, belongs to the group of four con-
nected to the i/o bus by the VX switches and, therefore, has a complete
charge path. All other capacitors are inactive, because one side of the
charge path is open. For illustration, the charge path of the capacitor cor-
responding to VX2 and VY4 is shown in heavy lines in Fig. 9 -10A.
27O TELEVISION BROADCASTING
The signals appearing on the VX and VY control buses are shown in

simplified form in Fig. 9 -11. With the exception of VX16, as explained
later, two positive pulses are applied from the X decoder to each VX con-
trol bus during each head scan interval. The first pulse connects the ad-
dressed capacitor to the i/o bus during the read time of the particular
Dual-Emitter Transistor Switch

VY1 VYf VY4
i
110 Bus 4i
ar
VX1
VX2 -1--
VX16
(A) Simplified schematic diagram.

110 Bus
X Drive
(B) Actual transistor switch.
Fig. 9 -10. VEC memory storage matrix.
scanning line, and the second pulse connects the same capacitor to the i/o
bus during the write time of the next scanning line. This is done to permit
predicting the slope of the ramp required for connecting the two MATC
steps.
ADVANCED COLOR-ERROR CORRECTION SYSTEMS 271
The pulses on the VX1 through VX15 buses form an overlapping, or

interlaced, sequence as shown in Fig. 9 -11. For example, the first VX pulse
to occur during line 2 of a head scan period is VX2, which occurs during
the read time of line 2; the second pulse generated is VX1, which occurs
during the write time of line 2. The pulses applied to VX16, however, do
not conform to this pattern because, in succeeding headwheel revolutions,
a seventeenth line occasionally occurs in each of the four head scan periods,
although only 16 capacitors are provided for each head scan. Since the volt-
age stored in the sixteenth capacitor must serve for both the sixteenth and
seventeenth lines, this capacitor can be written into only when a seven-
teenth line is present. This capacitor is always read out during the sixteenth
line of either a sixteen -or seventeen-line pass. In a 17 -line pass, however,
this capacitor must be read out again during the seventeenth line. The
ramps for the sixteenth and seventeenth lines, therefore, will be identical,
but little error will be introduced because the differences in slope between
adjacent ramps are normally small.
During a 16 -line head pass, the only pulse appearing on the VX16 bus
will occur during the read period of line 16. During a 17 -line pass, how-
ever, three pulses will appear on VX16, the first during the read period of
line 16, the second during the read period of line 17, and the third during
the write period of line 17.
During the scanning period of each of the four video heads, a train of
negative -going pulses is applied from the Y decoder to the corresponding
VY bus (Fig. 9 -11) . Each of these negative -going pulses corresponds to a
positive -going pulse on one of the sixteen VX buses. The corresponding
VX and VY pulses occur at the same time, thereby closing both the X and
Y switches of the addressed capacitor for the duration of the pulses.
Because the change in level between the first MATC step in a given head
scan and the last step in the preceding head scan does not represent a true
velocity-error difference, no capacitor is provided for storing this voltage.
Consequently, no write period exists during the first line. The first two
pulses in the VY train, therefore, occur during read periods, and the re-
maining pulses (except during a seventeenth line) occur alternately in
write and read periods.
9 -4. THE CHROMA AMPLITUDE CORRECTOR (CAC)
Like the VEC, the CAC consists of an analog system and a digital system.
The analog system compares the amplitude of the burst in the tape video
signal with a threshold level, generates an error voltage proportional to
the difference, and stores the error voltage in a capacitor memory. At the
beginning of each line, the voltage on the corresponding memory capacitor
is read out and used to generate two push -pull control voltages. These
voltages are fed to the fm equalizer of the tape machine, where they auto-
matically vary the equalization to keep the burst level constant.
Line 1 Line 2 I Line 3-{

Read _Write
VX1
-Read Write
VX2
Expanded View
Head Scan 1 Head Scan 2
Read' l Write
VX1 11
VX2 Il II
Il II
II 1
1 I I I
Il II
I II
ll II
II II
II I I
II I
II LL
II lI
II II
II II
VX16
I II III
Head Scan I
For 17 Line Pass Only
Vl'1
11IIIIII11111111III
Head Scan 2
VY2
1111111111111111111
_
VY3
11111111'
_
VY4 Sti
Illllf
Fig. 9 -11. Simplified timing diagram of VEC memory -control pulses.

Although the CAC error voltages are based entirely on the burst ampli-
tude, the burst actually constitutes a sample of the chroma signal. If there is
no differential gain in the system, the changes in equalization derived from
measuring the burst amplitude alone will be correct for the entire chroma
signal.
As in the VEC system, functions such as reading out of or writing into
the capacitor memory are controlled by semiconductor switches. These
switches are operated at the required times by pulses from the digital
system. Many of the digital pulses are identical to those used in the VEC
and are produced by the same circuits.
CAC Analog System

A functional block diagram of the CAC analog system is shown in Fig.
9 -12. (Basic waveforms are in Fig. 9 -13.) The system constitutes a closed
loop, since the burst input to the Cavec is obtained from the demodulator
of the machine, and the push-pull output signals of the CAC are fed back
to the fm equalizer in the machine. To simplify the discussion, however,
the operation will be described first as if the loop were opened at the point
marked "X" in Fig. 9 -12. Under these conditions, the CAC correction
voltages are not applied to the fm equalizer, and color errors due to chroma
variation therefore appear as differences in burst amplitude from line to
line.
Open -Loop Operation

In the machine, the demodulated video is fed to a filter circuit, which
removes spurious frequency components. The filter is installed in the burst
processor module during the CAC installation. The filtered burst is fed to
a half -wave detector and low -pass filter in the Cavec unit. This circuit re-
moves the subcarrier components and leaves only the half -wave rectified
envelope of the original burst. The envelope is applied to a peak detector
which produces a dc signal proportional to the peak amplitude of the
envelope. This signal is fed to a threshold detector which produces an
output proportional to the difference between the burst amplitude and a
reference level set by a potentiometer. This difference voltage, which con-
stitutes the CAC error signal, is fed to the i/o bus of the CAC memory
when the write switch is closed; it is then stored in one of the 64 memory
capacitors. The particular capacitor selected, or addressed, is determined by
the CAC digital system in a manner similar to that of the VEC digital
system. After the writing process has occurred, a clamp signal from the
digital system closes another switch which discharges a capacitor in the
peak- detector output to ground.
In this manner, there is stored in each of the 64 memory capacitors a
voltage which represents the average error between the burst and reference
levels for a particular line in a particular head scan interval. At the start
of each line, the CAC error signal stored in the corresponding capacitor is
-bc+i->
O _
_
o _
wá U
o
T
J
L
oce
U
._.
a
N
o o
706
LAI
Cr
y
v0
a U
Q
N
O _
m > >
E
m
vó
L)
in x
to
ai .
L p o
7
CO
7
Qm
Ñ O
r
U
L L
Fig. 9-12. Functional block diagram of CAC.

read out by applying it to the i/o bus. During the read interval, a switch
transfers the voltage from the i/o bus to a temporary- memory capacitor
in the sample- and -hold amplifier. The output of the sample-and -hold
amplifier is fed to two drivers which develop push -pull control voltages
(± V) proportional to the sampled error. These two signals constitute the
CAC control output of the Cavec unit.
Line A Line B Line C
Clock
red
Filteurst 11111
B
Half Wave
Detected
Burst
1II1 .11111 IIIII
Low-Pass
Filtered
Half Wave
Burst
Peak
Detected
Burst
Sample
Write
Clamp C
Peak
Detected
Burst
V
tV
Control
Fig. 9 -13. Basic waveforms in CAC.
Closed -Loop Operation

For closed -loop operation, the ± V signals from the Cavec are applied
to an electrically controlled variable attenuator installed in the fm equalizer
module during the CAC installation procedures (described in more detail
later) . This circuit performs the same function as that performed by the
manual equalization potentiometer before installation of the CAC. Changes

in attenuation vary the fm equalization by changing the relationship be-
tween the energy in the sidebands and the energy in the carrier. The
phasing of the ± V signals is such that they tend to reduce variations in
burst amplitude. Thus, the system acts as a feedback loop, with the ± V
signals providing the negative feedback. Because of the 64 memory capaci-
tors, the system has the same effect as 64 closed loops, one for each line in
each head scan.
To permit effective closed -loop operation, the read -out and write -in of
the CAC error for a particular line must occur during that same line, and
read-out must occur before the burst appears at the fm input of the fm
equalizer. As a result, the read -out and the resulting correction occur
shortly after the leading edge of the tape horizontal signal, as shown in
Figs. 9 -13 and 9 -14.
For a 17 -line head pass, on 525 -line standards, the error voltage on the
capacitor written into during the sixteenth line is read out during both the
sixteenth and seventeenth lines. (Similarly, on 625 -line standards, the
capacitor written into during the fifteenth line is read out during both the
fifteenth and sixteenth lines.) The inaccuracy thus produced is not visible
Burst Interval
FM Input to Equalizer
Equalizer Push -Pull

Control Signals
CAC Corrected Burst
Demod Video Output

Sync Tip
Fig. 9 -14. Timing relationship of bursts in fm and video.

in the picture because a seventeenth line does not occur periodically during
successive scanning periods of the same video head.
CAC Digital System

As in the VEC, previously described, the basic timing signals for the
CAC are derived from the tape -horizontal, 2 X 1 switcher, or 4 X 2 switcher
signals. In fact, most of these signals are produced by the same circuits
for both the CAC and the VEC.
The CAC uses the same line counter as the VEC (Fig. 9-15 ) The .
counter produces 16 unique binary combinations of output signals, one

for each line of a 16 -line head scan. (When a seventeenth line is present,
it is detected by additional logic circuits.) The counter output combinations
are decoded to produce sixteen X address signals for the capacitor memory;
these are referred to as the CX signals to distinguish them from the VX
signals used in the VEC.
A CY strobe signal (similar but not identical to the VY strobe) is
generated by combining the read and write pulses for each scanning line.
This signal and the 2 X 1 and 4 X 2 switcher signals are combined in a
decoder to produce the four Y address signals for the memory, CY1
through CY4. Each of these signals consists of a train of 16 or 17 pairs of
read and write pulses (15 or 16 for 625 -line standards) occurring only
during a particular one of the four head scanning intervals. The information
for determining the head scanning intervals is obtained from the 4 X 2
and 2 X 1 switcher signals.
As in the VEC, each pair consisting of one CX signal and one CY signal
addresses a particular capacitor in the CAC memory. Thus, through the
use of a switch matrix controlled by the CX and CY signals, the capacitors
can be addressed sequentially. When the pair of signals corresponding to a
given capacitor occurs simultaneously, one side of the capacitor is con-
nected to the i/o bus, and the other is connected to ground, thereby com-
pleting the charge path of this capacitor only. Unlike the case of the VEC,
the same capacitor is addressed during the read and write intervals of a
given scanning line.
Electronically Controlled FM Equalization

In an fm equalizer module unmodified for Cavec, equalization is accom-
plished by delaying the fm signal and then subtracting the delayed signal
from the undelayed signal in a differential amplifier. Both the crossover
frequency of the equalization curve and the degree of equalization (de-
parture from flat response) can be adjusted. The crossover frequency is
adjustable in steps by a five -position switch that selects different amounts
of delay. The degree of equalization is adjustable by a front -panel po-
tentiometer which controls the amplitude of the delayed signal. (The larger
the signal in the delayed path, the less the equalization is, and the smaller
the chroma content of the video signal is.)
Fig. 9 -15. Simplified diagram of CAC memory -control circuits.

When the fm equalizer is modified for CAC, a circuit board containing

an electrically controlled attenuator is added to the module (Fig. 9-16 ) .
This circuit is inserted in the path of the delayed signal in place of the
potentiometer. The ± V signals from the CAC output of the Cavec are
then used to control the attenuation. When the manual /auto relay ( Fig.
9 -17) is in the auto condition (energized) , the ± V signals are derived
from the error signals stored in the CAC memory. The equalization is thus
controlled automatically on a line -by -line basis. Under these conditions,
the manual equalization control on the equalizer module is disconnected,
and the only way that the degree of equalization can be changed manually
is by adjusting the BURST RATIO control on the front panel of the Cavec
unit. When, however, the relay is in the manual condition (de- energized) ,
the ± V signals are derived from a dc voltage which depends on the setting
of the manual equalization potentiometer on the fm equalizer. The equali-
zation can then be controlled manually by adjusting this potentiometer.
Fig. 9 -18 is a schematic diagram of the fm attenuator board. The circuit
consists of diode quad Ml, transistors Q1 and Q2, and two regulated sup-
plies consisting of R4, X1, and Cl and R7, X2, and C2.
Direct
Signal Path
Delay Line Equalizer Difference Equalized
FM Input-N. Driver Delay Line Amplifier FM Output
Reflected
Signal Path
Diode Quad
FM In (Attenuator) FM Out
+V Signal
Cavec Curren
(CAC) Source
-V Signal:
FM Attenuator Board
(A) Block diagram.

2 Min Eq Min Attenuation
Max Eq (Peaked Video)

Max Attenuation
f = FM Crossover
° Frequency
2f0
0
FM Frequency (Input)
(B) Equalization curves.

Fig. 9 -16. Principle of electrically controlled fm equalization.
An output voltage of -3.9 volts is provided by R7, X2 and C2; R4, X1,
and Cl provide +3.9 volts. Transistor Q2 (npn) diode quad Ml, and ,
transistor Q1 (pnp) are connected in series between the -3.9- and +3.9-
volt supplies. The delayed fm signal is fed into junction A of Ml, and the
output is taken from junction B. The +V signal from the Cavec module is
applied to the base of Q1, and the -V signal is applied to the base of Q2.
The networks containing R1, C3, R3, and C4 filter the ±V signals. The +V
r
Direct
FM Input-LP. Driver Equalizer

Delay Line -
Signal Path
r
.
Difference
Amplifier
--FM Output
Equalizer I 2 3 4 5
Step Select °o-

o----
Reflected q Reflected
Signal Path Signal Path
r
Driver Diode Quad
FM Attenuator
Manual Board
EqControl
(On Eq Module -3.9 V
Front Panel) Current
Source
FM Equalizer
+V Control -V Control
Manual
o .
Push-Pull
Driver
Circuits
CAC Error 1 o f:
Manual IOff l
Auto
CAC Man uall Cavec f Play ILS /BS

Auto Relay Front Panel Bus
Auto (On)
26V Switch
Cavec Unit CAC
Fig. 9 -17. Partial block diagram of fm- equalizer control system.
and -V signals control the current through the path consisting of Q1,
junctions C and D of M1, and Q2. The current decreases as the +V signal
assumes more positive values ( and the -V signal assumes more negative
values) because the transistors approach cutoff. For fixed values of +V
and -V, the current is constant.
Since the four diodes are forward biased, the fm signal applied to junc-
tion A can flow through the quad to the output at junction B. The im-
pedance between junctions A and B, however, depends on the amount of
forward bias current. This, in turn, depends on the current supplied by QI
and Q2. The quad, therefore, attenuates the fm input by an amount depend-
ing on the ±V signals. The larger the +V signal, the greater is the
attenuation of the delayed -path fm signal and the greater is the chroma
content of the video. Conversely, the smaller the +V signal, the smaller are
the attenuation and the chroma content of the video. Use of the two bal-
anced push -pull drive signals reduces transient disturbances and distortion
of the fm signal.
+12V
+3.9V
C1y
+V Control
From CAC
I
FM Input FM Output 1
R2 R8
-V Control
From CAC
3 C4
For Manual
3.9 V Equalization
Control I n
FM Equalizer
-12V
Fig. 9 -18. Simplified schematic diagram of fm- equalizer attenuator board.
The -3.9 -volt zener regulator circuit (R7, X2, and C2) , in addition to
providing bias to Q2, provides the input voltage to the manual equalization
potentiometer on the fm equalizer module. The arm of this potentiometer
is connected through a resistor divider to the manual /auto relay in the
Cavec unit. When the relay is de- energized, the attenuated potentiometer
voltage is fed to the input of the first push -pull driver stage. Consequently,
since the ±V signals depend on the potentiometer voltage, the equalization
can then be varied manually by adjusting the potentiometer. When, how-
ever, the relay is energized, the potentiometer voltage is disconnected, and
the CAC error voltage from the sample- and -hold capacitor is fed to the first
push-pull driver. Under these conditions, the potentiometer has no effect,
and the equalization is controlled automatically by the CAC signals from
the capacitor memory.
EXERCISES
Q9 -1. Give the two primary causes of color banding.

Q9 -2. What are the primary causes of (A) hue banding and (B) saturation
banding?
Q9 -3. To what value is any single line of color (hue and saturation) cor-
rected?
Q9 -4. What is the basic function of (A) the analog circuits and (B) the
digital circuits in Cavec?
Q9 -5. Why doesn't the Automatic Time Correction (ATC) circuitry remove
color errors occurring in each head pass?
Q9 -6. What does VEC basically do to correct velocity errors?
Q9 -7. Flow are the correct slope and amplitude of the ramp described in
A9 -6 obtained?
Q9 -8. Why is the particular VEC memory capacitor that is addressed during
readout not written into during that line, but written into during the
following line (Fig. 9 -11) ?
Q9 -9. In Fig. 9 -10, what are the X and Y signals of the VEC memory -control
pulses?
Q9 -10. Do the CAC memory capacitors have the same readout and write -in
timing as the VEC memory capacitors?
CHAPTER 10
Splicing and Editing for

Quadruplex Recording
Motion -picture film can be examined on a frame -by -frame basis at zero
speed for the purpose of cutting and splicing for editing. Video tape, how-
ever, produces no output unless it is moving, and it must be moving near
the recorded speed for a usable picture. Thus video -tape editing is essen-
tially a dynamic function rather than a static one, and it requires special
considerations.
It is natural that the first approach to video -tape editing or splicing was
a mechanical operation which involved making the magnetic tracks and
editing marks visible ( see Fig. 2 -10 in Chapter 2 for an example) so that
tapes could be cut at the proper place to avoid picture disturbances. Me-
chanical splicing is still used in limited applications, and we will consider
this technique first. We will then examine the more recent systems of elec-
tronic editing, and, finally, we will study automatic (computerized) editing
programmers.
First, it should be known that early tape systems (monochrome) did not
operate synchronously with local station sync, and 60 edit pulses per second
(one pulse per field) seemed most appropriate. However, when synchron-
ous operation became a practical reality, a specific field was used for refer-
ence so that splicing always occurred on the same field to avoid a momentary
half -line 'bobble." Therefore, the edit -pulse rate became 30 pulses per
second, and these pulses are termed frame edit pulses. Later, with the ad-
vent of color, the edit-pulse rate became 15 pulses per second in many
systems so that the four -field cycle of the NTSC color signal could be
accommodated.
10-1. MECHANICAL SPLICING

Fig. 10 -1 illustrates the steps in splicing video tape using the RCA
optical tape splicer. A 40 -power microscope is employed for location and
283
precise alignment of frame pulses. To eliminate rollover and /or temporary

horizontal instability following a splice, the cut must be properly framed
with the edit pulse and formed to a clean, solid, square butt-splice.
A mechanically perfect video -tape splice can be accomplished by observ-
ing precautions of cleanliness, application, and handling. Dust, dirt, debris,
fingerprints, and solvents must all be carefully removed prior to the applica-
tion of the splicing tape. The adhesive area of the splicing tape should not
(A) Stop recorder at desired location (B) Develop tape with solution of
and cut tape. iron oxide to locate nearest edit or
frame pulse.
(C) Align edge of video track on (D) Bring cutter into position and
built -in reticle to cut tape halfway be- make cut.
tween adjacent video tracks.
Fig. 10 -1. Steps in mechanical

be touched or exposed to airborne contamination by excessive exposure to

air before application. Adequate pressure must be applied to the splice to
remove air bubbles and to assure proper sealing. With the use of a quality
splicing tape and proper splicing techniques, a video -tape splice will last
indefinitely, therefore contributing additional tape life and economy.
There are three basic reasons why a splice fails: improper application,
contamination, and improper handling of the splicing tape. Any one of
(E) Drape tape from supply reel over (F) Align tape track prior to
tape support post and position under making cut.
left hold -down door.
(G) Shear tape to be joined to tape (H) Dispense splicing tape and
already in splicer. prepare to put it in place.
splicing of video tape. (Cont'd on next page.)

these elements or a combination of them can cause splice failure and exces-
sive tape and head wear.
The adhesive used in a splicing tape is pressure sensitive. This special
adhesive, to assure a perfect bond, requires adequate pressure during appli-
cation. Occasionally, an operator does not apply sufficient pressure to the
entire surface of the splicing tape, causing an improper bond between the
video tape and the splicing tape. After the splicing tape is applied to the
(I) Insert splicing tape beneath (I) Press both ends of the tape firmly
the lifted ends of the two tapes to provide a perfect butt joint.
to be joined.
(K) Trim away excess splicing tape (L) Examine the completed splice.
with trimming tools.
Courtesy RCA
Fig. 10 -1. Steps in mechanical splicing of video tape. -cont'd.
video tape, pressure may be applied across the entire surface of the splicing
tape with a clean blunt instrument or clean rubber roller.
Proper alignment and centering of the splicing tape directly over and
in line with the splice cut must be assured for maximum strength. If the
splicing tape is not centered over the spliced sections, a lopsided splice
will result, with the narrow side being weak and susceptible to failure. If
any of the splicing -tape edges are not tightly sealed to the video -tape
backing, contamination will tend to collect at the areas of exposed adhesive.
Contamination deposits may cause tape damage and head wear in a manner
similar to the action of air bubbles.
Contamination of video-tape splices is caused by a variety of sources such
as solvents used for cleaning the tape, fingerprints, debris from transports
and splicing fixtures, and airborne dust.
Weakening and subsequent failure of splices can occur if the solvent
used to clean the video tape ( such as that used to remove splicing powder)
is permitted to seep onto the adhesive of the splicing tape. The finite
separation which exists between the sections of video tape being spliced can
allow the solvent to deteriorate the adhesive. Most commonly used solvents
(alcohol, Freon TF') are low -viscosity solutions which can easily penetrate
the small splice separation. To avoid solvent damage to the adhesive, be
certain that the tape is completely dry before applying the splicing tape.
Fingerprints, debris, and wear products can weaken the bond between
the splicing tape and the video -tape backing. A fingerprint is a deposit of
body oil which greatly reduces the bonding strength of the pressure- sensi-
tive adhesive. Fingerprints on either the video tape or the adhesive surface
of the splicing tape can be eliminated by proper handling or by wearing
lint -free gloves.
Ideally, the video -tape backing should be cleaned with a cotton swab
moistened with Freon TF to remove fingerprints and contamination. After
cleaning, the splice area should not be touched with exposed fingers. When
the solvent has evaporated and the tape is completely dry, the video -tape
sections can be aligned and the splicing tape applied and secured.
Cleanliness of the splicing fixture and the immediate work area must
be conscientiously observed during splicing operations. The knife portion
of the splicer and the splicing -tape application plate must be kept meticu-
lously clean to avoid trapping contamination within the splice. In addition,
the loose end of the splicing tape which protrudes from the dispenser is
subject to airborne contaminants and loose debris that can make it unsuit-
able for use.
In handling the splicing tape, avoid touching the adhesive surface which
will contact the video -tape backing, protect the splicing tape from con-
tamination, and do not distort or stretch the splicing tape during
application.
'DuPont Trademark
10 -2. ELECTRONIC SPLICING (AMPEX)
With electronic splicing, the recorder may be started and stopped at

random between scenes for costume or scenery changes, for animation
effects, or for any other special purpose. In addition, the recorder may be
used to insert new scenes or commercials at any chosen point in the middle
of a previously recorded program. This feature permits the correction of
production "fluffs" that may have occurred anywhere in an otherwise
perfect recording.
The heart of the Ampex system is the Electronic Editor, a solid -state
unit that modifies the switching logic of a standard Ampex television
recorder. Consider the situation that would exist in an unmodified recorder
if a recorded tape were threaded onto the machine, and, at some point
during the replay of this tape, the record button were operated to record
a new scene into the middle of the existing program. Record current would
be fed to the video record heads to record the new scene, and erase current
would be fed to the erase head to erase the old scene. However, a number
of problems would arise which would make the attempted transition from
the original recording to the new recording impossible to achieve. These
problems involve servomechanism disturbances, incorrect phasing of the
reproduced tape signal and incoming video signal, and an overlap of
recording caused by the distance between the video record and erase heads.
Fortunately there are straightforward solutions to each of these problems.
It is possible to arrange the servo system so that it uses the same sources of
drum and capstan drive signals on entering the record mode that it uses
during the play mode. Specifically, the drum drive signal remains phase
locked to the system vertical sync, and the capstan drive source remains a
60 -Hz Wien -bridge oscillator. During the record mode, this oscillator is
free -running, but no difficulty from long -term drift occurs, because any
minor variation in frequency will be tracked out during replay.
The correct phasing of the reproduced tape signal to the incoming video
is obtained by using the Intersync television signal synchronizer. This unit
is a precision servo -control mechanism that guarantees that synchronizing
pulses recovered from the tape have a very close time relationship to
synchronizing pulses from the studio sync generator.
The problem of the distance between the video record and erase heads
remains. This distance (Fig. 10 -2A) would cause a double recording to
exist for 9 inches of the tape. The double recording, in turn, would cause
both the sync -processing and servomechanism circuits to malfunction. To
deal with this problem, the distance must first be translated into units pre-
cisely and readily measurable by electronic means. It has been found con-
venient to do this by considering the distance as a time interval correspond-
ing, at the linear tape speed of 15 in /s, to slightly less than 18 television
frames. On the section of tape in Fig. 10 -2B, line S is exactly 18 frames
ahead of the video record head and is slightly ahead of the intended splice.
Video Record Heads
Video Erase Head Tape Travel-ow
Approx 9 Inches
(A) Separation of erase and record heads.

Video Record Heads
Video Erase Head
Tape Travel -0. A
J
1/1111111111/1/1
Line S - 18 TV Frames
(B) Measuring head separation.

v Frame Intervals
Line S Vertical -Sync Pulses

Tape Travel
Guard Bands Video Tracks

(Width-5Mils) (Width -10 Mils)
(C) Detail of splicing point.
-; 1-8 ms
Video Record Heads
i' Video Erase Head Tape Travel -- A
'l
i 111111
Frame Intervals
Line S
18 Ti Frame
(D) Initial position of splicing point.

Fig. 10 -2. Head separation and location of splicing point.
This line is further defined as existing in the guard band following a video
track that bears vertical- synchronizing pulses, as shown in Fig. 10 -2C.
In the operation of the Electronic Editor, the location of line S is deter-
mined after a random cue by means of a vertical -gating circuit, which
detects the first frame synchronizing pulse following the cue. At this
instant in time, the tape lies under the heads as shown in Fig. 10 -2D. The
times taken for line S to travel to the video erase head and the video record
heads may be defined as 8 milliseconds and 18 frames, respectively.
The 8- millisecond delay is obtained from a delay multivibrator; at the
end of this period, the erase head is energized by an electronic gating
circuit. Turn -on time is approximately 30 microseconds, and erasure starts
exactly on line S (Fig. 10 -3) The 18 -frame delay is computed by a binary
.
counter system. At the end of this period, the record current is gated on
electronically. Turn -on time in this case is 2 microseconds, and the new
recording starts immediately following line S, as shown in Fig. 10 -4.
Tape Travel -a Line S Vertical -Sync Pulses
Erased Tape Original Recording
Start of Erasure

Fig. 10 -3. Start of video erasure.
Tape Travel - - Line S
New Recording Erased Tape
Start of New Recording

Fig. 10 -4. Start of video recording.
The accuracy obtained from the foregoing system is approximately 0.05

percent, which enables the operator to make perfect butt splices by merely
pressing the record button. The preservation of all video information at
the splice and the accuracy of tracking is evident from an examination of
Fig. 10 -5.
The Electronic Editor has two modes of operation, referred to as insert
and assembly. In the assembly mode, the unit may be used to add further
program material to the end of an existing program, and, in this way,
assemble a complete program from individual short scenes. In fact, a
completely edited tape may be prepared with only one camera and one
recorder.
The insert mode is used to insert fresh material into the middle of an
existing tape. In this mode, commercials, for example, may be added to a
recorded program at any time following its original preparation. Also,
because any information behind the inserted material is erased and there is
no loss of sync, it is possible to correct errors in the original production.
If the action is repeated and the RECORD button is pressed, the scene con-
taining the production error is replaced with a new scene.
Obviously, in the insert mode it is necessary to make two splices per
operation and still maintain synchronism. The first (or ingoing) splice
was described previously. The second (or outgoing) splice is made in a
similar manner, except that at the end of the measured time intervals, the
erase and record currents are turned off instead of on. This is performed
with the same precision as that of the first splice.
Tape Travel -I. Line S
New Recording Over Tape

Erased by Video Erase Head
11111 Original Recording
Splice

Fig. 10 -5. Electronic -splice configuration.
During the recording of the insert, the capstan -oscillator frequency is not
controlled; therefore, it may drift slightly if the inserted material is of
considerable length. This drift, in turn, will cause a variation in the wave-
length of the recorded control -track signal and result in an abrupt phase
shift in the reproduced control -track signal at the outgoing splice.
One solution to this problem would be to employ a phase- correction
system in the capstan -oscillator output circuit during recording. This, how-
ever, would be extremely complex. The adopted solution utilizes the fact
that in the insert mode a control -track signal already exists on the tape.
This information would normally be erased by the full-width video erase
head; however, a new head which has a separate section for the control -
track erasure has been developed (Fig. 10 -6). The control -track section is
disconnected when the insert mode is selected, and, although the machine
is placed in the record mode, reproduction of the control track continues.
Thus, it is possible to retain normal control of the capstan oscillator, and
any possibility of control -track phase shift is avoided.
33' (Exaggeratedl-
Audio Track Area
Video Erase Head
Fig. 10 -6. Detail of erase -

head assembly.
Cue Track Area
Control -Track
Erase Head
The video erase head is a new design which gives an improved erase
efficiency. It uses only a half turn in contact with the tape, and the elec-
trical gap width is 5 mils, the same as that of a guard band. The gap is
optically straight and inclined to the perpendicular at an angle of 33
minutes of arc. Thus, the erasure pattern is exactly parallel with the video
record pattern. The audio and cue tracks are not affected by the new head,
and control -track erasure is optional.
It will be seen that the functions of the Electronic Editor are of the logic
type and serve to provide precisely timed and electronically gated video
erase and record currents. A block diagram is given in Fig. 10 -7. The
various delay multivibrators, flip-flops, gates, and binary counters are all
classic circuit configurations that will not be discussed at this time. The
overall functions, however, may be traced as follows.
Following the initiation of the record mode, a delay multivibrator pro-
vides a 60-millisecond delay during which all normal record relays have
time to operate. Flip -flop 1 then operates and places AND gates 1 and 3
in a ready state. Pulses, derived from the Intersync unit, that mark the third
vertical-sync pulse in every frame interval are reshaped in a pulse former
and routed to AND gates 1 and 2. Gate 1, which is in a ready state, there-
To
Record
Relays
Record
1 Flip -Flop 1 AND Gate 1
Variable
One -Shot 1
Binary
Counters
AND Gate 3
8 ms
Frame
Pulses Variable
One- SOW 2 Flip lop 3 ---tow Erase
Switch
Flip-Flop 4 RF Switch ---rlo
570 us Record
Driver
To
Erase
1 Stop
Flip -Flop 2 AND Gate 2

Variable
One -Shot 3 -O. AND Gate 4
Amp
Pros
Fig. 10 -7. Block diagram of Ampex Electronic Editor.
fore operates at the first frame pulse and triggers variable delay one -shots
(multivibrators) 1 and 2. One-shot 1 serves to provide the necessary 8
milliseconds of delay before the video erase current is turned on. Erasure
turn -on is acomplished by flip-flop 3 and an electronic switch.
Because timing is referenced to the third pulse in the vertical-sync train,
and at this time the rotating video -head drum is positioned to place the
active head tip at the center of a video track, a 570- microsecond delay is
provided by one -shot 2. This delay allows the head to travel to the end of
the track, at which time the succeeding head on the drum periphery is
positioned at the beginning of the next track. From this timing reference,
18 television frames are counted off by the binary counter system, and at
the end of this period gate 3 operates. Video record current is then turned
on by the action of flip -flop 4 and a second electronic switch.
Timing for the outgoing splice is provided by the operation of gate 2,
gate 4, and variable delay one -shot 3. The operation of these stages, to-
gether with the reuse of the counter train, resets flip -flops 3 and 4 and
places the two electronic switches in the off condition.
An additional flip-flop (not shown in the block diagram) is operated
by the binary counters exactly 18 frames after the video record current is
turned off. This permits the audio track to be cleared of all extraneous video
information by the action of the audio erase head. (New audio information
has, of course, accompanied the new video information.) A time -delay
circuit allows all audio functions to revert to normal before the machine
is stopped. In this way, any possibility of transients on the audio track is
avoided.
The quality of the finished splice can be observed in Fig. 10 -8, which
shows the video, blanking, and synchronizing information in the video
waveform at the moment the electronic splice passes the video heads. The
splice appears the same as a change of picture caused by camera switching.
The rapidity with which the editing function is performed becomes
apparent when it is realized that the entire splicing operation is performed
while the recording medium is in motion at normal speed.
Fig. 10 -8. Vertical waveform at

moment of splice.
10 -3. ELECTRONIC SPLICING (RCA)

The RCA Electronic Splicing Accessory permits recording a new video
signal on a tape containing a previously recorded signal so that a smooth
transition, similar to a video switch transfer, will occur during playback.
The equipment is designed to operate on 50- or 60 -field standards and at
normal or half tape speed.
Types of Splices
In an electronic splicing operation, the junction of the two video signals
at the beginning of the new recording is called an ingoing electronic
splice. If a portion of the old recording is retained after the new section,
the junction at the end of the new recording is called an outgoing electronic
splice.
Typical ingoing and outgoing splices are shown in Fig. 10 -9. Note that
the new video starts or ends near the top of a track containing vertical
sync. Also observe that a small unerased section, on which the new record-
ing is superimposed over the old, exists in both splices. The overlaps are
deliberately created to allow for differences between the machines on which
the two recordings are made and will produce no subjective disturbance
except in animated recordings. However, since the series of splices in an
animated recording are made on the same machine, no tolerances are in-
volved, and the overlap can be eliminated by an adjustment.
The overlap for an ingoing splice is normally adjusted for 11/2 tracks
with the supply reel fully loaded and the tape running at normal speed.
The splice timing circuits then automatically change the overlap to 21/2
tracks for an outgoing splice. When the tape is running at half speed, the
overlap is automatically doubled. Thus, with the preceding adjustment,
the overlaps become three and five tracks.
New Video Recording New Video Recording

Video Tracks Starts Here Stops Here
Video Erasure End of Frame Vertical Blanking

Starts Here End of Frame Vertical Blanking Video Erasure
Stops Here
(A) Ingoing, (B) Outgoing.

Fig. 10 -9. Typical electronic splices.
Formation of Splices
An ingoing splice is formed by applying the video erase and record
currents to the respective heads at different times, instead of at the same
time as in normal recording. This is necessary because the master erase
head (in the RCA system) is approximately 7.5 inches from the headwheel,
and, if the two currents were applied simultaneously, the two signals would
overlap, without erasure, on the entire length of tape between the heads.
Similarly, formation of an outgoing splice requires that the video erase
and record currents be shut off at different times since, if they were shut
off at the same time, the section between the heads would be erased but
not rerecorded.
The timing of the erase and record functions during formation of a
splice is automatically determined by an electronic counter which, in effect,
counts the recorded frames on the tape as they pass the headwheel (Fig.
10 -10) The counter controls separate delay circuits for recording and
.
erasing. In turn, the delay circuits control transistor switches that either
allow or prevent application of the currents to the heads.
Add -On and Insert Operation

Two types of electronic splicing, called add -on and insert recording, are
possible. Each type requires a different set of operating conditions.
c
Q
~ T rT
zo
TO
O E ç
G-
-d ú
Uo -
ó u
U O Q 2
L
An
° ú
U
o ac
v ou L
"En
E o
n ó >
m ce
a a LN -
Ln Ó r_
Ñ
$ a
o Ñ 3
an 3 ò
Q iJ
O 2 l Z.
U N Á
u
a c
jE
TO
-r vNi eeT O Of
U c
=o
.
WL N N Ñ
>
O
Ó
~Ú
U
a O E
-
.U
ó
N
N
N
ce - -43
Q Z
Fig. 10 -10. Simplified diagram showing principles of electronic splicing.

In an add -on operation ( also called assemble) , the new recording is

merely added at the end of the old recording, and therefore no outgoing
splice is required. A control track is recorded as in normal operation, but
the capstan servo is modified by use of a capstan memory circuit which
matches the speed of the new recording to the original tape speed. The
capstan memory circuit provides a steady dc voltage which serves as a
reference for the capstan oscillator during recording. The dc voltage is
related to the original tape speed by adjusting a potentiometer while play-
ing back the tape before recording.
In an insert operation, the new recording starts at any desired place but
ends before the old recording ends. Therefore, since part of the old record-
ing is retained, an outgoing splice is made in addition to an ingoing splice.
Also, the original control track, which already exists on the entire length
of the tape, is retained and played back to control tape speed during the
recording. Retention of the control track is made possible by replacing
the original master erase head with a selective erase head containing
separate sections for the control and video tracks. A relay removes the
erase current from the control -track section during an insert recording.
Mode Selection
A selector switch on the splice logic module permits placing the machine
in the add -on, insert, or normal mode. As a special feature of the equip-
ment, the selector switch may be left in either the Add -On or Insert posi-
tion, and the machine will still make a normal recording if the RECORD
button is pressed before the PLAY button. This feature is obtained by use
of a splice -mode circuit on the splice control module; the circuit is similar
to the mode memory circuits of the control system.
When the switch is on Add -On or Insert and the PLAY button is pressed,
the splice -mode circuit becomes energized, and the machine enters the
splice play mode. In this mode, the splice circuits are placed in a state of
readiness for insert or add -on operation, as selected, and the machine will
not go from play to record unless the servo selector switch on the tape sync
processor module is on Switchlock and the machine is actually in Switch-
lock. In addition, the following operation and setup controls are actuated:
1. Aud Record Off position of toggle switch on record control panel.
2. CAP ADJ /FIXED TW /VAR TW switch and CAPST ADD ON ADJUST
control on splice logic module.
3. SPLICE RECORD HOR PHASE control on splice timing module. (If
machine is not in Pixlock and switch on splice logic module is on
Var TW.)
4. ERASE TEST buttons on splice timing module.
If the switch is on Add -On or Insert and the audio or cue record buttons
are pressed, the video circuits will enter the play condition. Therefore, the
splice -mode circuit will be energized, and, except for the audio or cue
record circuits, the machine will enter the splice play mode. When the
switch is in the Norm position, the machine cannot enter a splice mode
and therefore functions normally. Similarly, if any of the splice modules
are removed, a splice interlock circuit is broken, and the machine reverts
to normal operation.
Electronic splicing is facilitated by installation of a new ADD ON /INSERT
indicator and modification of the original SL ( Switchlock) indicator. If the
splice selector switch is placed on Add On or Insert but the splice -mode
circuit is off, the corresponding section of the ADD ON /INSERT indicator
glows red. However, when the splice -mode circuit goes on, the lamp color
changes to white. Similarly, when the servo selector switch on the tape
sync processor is on SL, the PLAY button is pressed, but the machine is not
yet in Switchlock, the SL indicator glows red; when the machine actually
enters Switchlock, the indicator color changes to white.
Provisions for Retaining Original Audio and Cue Tracks

To permit retaining the original audio or cue track, if desired, while
recording new video tracks, a number of modifications are made to the
record control panel during installation of the splice equipment. These
modifications consist of replacing the original audio and cue release push
buttons with three -position toggle switches, rewiring the audio and cue
mode indicators to provide either on or off indications, and installing a
transistor switch, SA14A, on the mounting bracket of the splice control
module, behind the panel.
On each toggle switch, two of the positions, marked Off and Norm, are
fixed, and the third, marked Release, is momentary. When both switches
are in the Norm position, the corresponding sections of SA14A connect
the master record bus to the audio and cue record buses, as in the original
machine. Therefore, when the MASTER RECORD button is pressed, recording
will occur on the audio, cue, and video tracks, and the on sections of the
audio and cue record indicators will light.
When the cue toggle switch is in the off position, it turns off the corre-
sponding section of SA14A, applies -26 volts to the unlatch input of the
cue mode module, and lights the off section of the cue record indicator.
Therefore, recording will not take place on the cue track if the MASTER
RECORD or CUE RECORD buttons are pressed.
If the audio toggle switch is placed in the off position, the results depend
on whether the splice-mode bus is energized. If the splice bus is energized,
the corresponding section of SA14A is turned off, -26 volts is applied to
the unlatch input of the audio mode module, and the off section of the
audio record indicator is lighted. Consequently, the audio recording circuits
cannot be energized. If the splice bus is not energized, the same actions
occur, except that SA14A is not turned off. Under these conditions, the
audio recording circuits will be energized if the MASTER RECORD button
is pressed, but not if the AUDIO RECORD button is pressed.
The momentary release positions of the toggle switches serve the same
functions as the original momentary push buttons. If the machine is in
audio or cue record, or both, and either switch is moved to the release
position, the corresponding record function will be turned off.
Horizontal- Phasing Provisions
To prevent a horizontal lurch in the picture, the phasing of the tonewheel
with respect to the headwheel during a splice recording must match that of
the original recording machine. Therefore, to permit adjusting the steady -
state headwheel phasing, a ten -turn potentiometer marked SPLICE RECORD
HOR PHASE is provided on the front panel of the splice logic module. The
control operates in conjunction with the last two positions of the CAP ADJ/
FIXED TW /VAR Tw toggle switch on the splice timing module. When
this switch is in the Fixed TW position, the phasing potentiometer is
disabled. If the toggle switch is on Var TW, the phasing potentiometer
will be actuated when the machine enters the splice play mode. The control
may then be adjusted as directed in Chapter 11 to make the new recorded
sync occur in the same places on the tape tracks as the original recorded
sync.
Control of Tape Speed During Add -On Recording

As previously mentioned, during an add -on recording the capstan oscil-
lator is referenced to a dc voltage obtained from a capstan memory circuit.
A control, marked CAPST ADD ON ADJ, on the splice logic module permits
adjusting this voltage for the correct tape speed when the machine is in
the splice play mode and the CAP ADJ / FIXED TW/VAR TW toggle switch is
held in the momentary Cap Adj position.
Erase Test Feature

An erase test circuit is provided to permit adjusting the video erase
timing without the need for making the recorded tracks visible with a
developing solution. The operating controls for this circuit consist of two
red buttons marked ERASE TEST and a potentiometer marked ERASE DELAY
on the front panel of the splice timing module. The two buttons are
momentary and are connected in series. This arrangement prevents unde-
sired erasure of the tape in case the operator presses one of the buttons
accidentally.
When the machine is in the splice play mode and the two buttons are
held down, the erase test circuit causes the tape to be erased in pulses of 200
microseconds duration. The timing of these pulses with respect to the
recorded sync is the same as if a series of ingoing splices were being made.
The effect of this pulsed erasure is visible on an oscilloscope connected to
the fm switcher output as a "butterfly" interruption of the signal. Since the
butterfly position depends on the setting of the ERASE DELAY control, the
operator can easily adjust the control to start erasure at the desired time.
This test should be made only while playing back a tape containing
unneeded material, since the tape will be erased. Once the adjustment is
made, it will require no further attention until the headwheel is changed.
A short functional description of the electronic splicing system is given
in the following paragraphs. Fig. 10 -11 is a functional block diagram of the
system, and Table 10 -1 lists the major circuit designations and functions.
Add -On Operation

For add -on operation, the mode selector switch on the splice logic
module is placed in Add On, and the tape is played back until the place
where the splice is to be made is found. The tape is then rewound and the
PLAY button is pressed. This action turns on the splice mode circuit on
the splice control module, which energizes the splice bus. The splice bus
then turns off the normalizer circuit on the splice timing module, which
allows the machine to enter the splice play mode. In this mode, the Switch-
lock detector on the splice logic module applies an inhibit voltage to the
master -record mode module (unless the machine is in Switchlock) ; the
reference generator, capstan oscillator, and guide servo are locked in the
play condition; and the splice switches on the erase oscillator module and
auxiliary circuit boards are turned on.
The operator then observes the Switchlock indicator to see whether the
machine has entered Switchlock and, if it has, presses the MASTER RECORD
button approximately 1/2 second before the predetermined splice point if
the tape is running at normal speed, or one second before if the tape is
running at half speed. This action turns on the master -record mode module,
and the output bus of this module starts the electronic counter in the splice
timing module.
At this time, the various record and erase functions are inhibited by the
splice switches and logic circuits, although the master record bus is ener-
gized. However, as the recorded section then present between the master
erase head and the headwheel moves past the headwheel, an automatic
cycle timed by the counter occurs, during which the inhibitions are re-
moved at the proper times for an ingoing splice.
During the cycle, the counter is timed by reference frame pulses derived
from external sync. Since the machine was previously in Switchlock, each
of these pulses occurs at the same time that the headwheel is scanning a
track containing vertical sync at the end of a frame. Therefore, in effect,
the recorded frames on the tape are being counted.
The timing of the splice cycle depends on the number of recorded
frames between the master erase head and the video headwheel. This num-
ber, in turn, depends on the distance between the heads, the tv frame rate,
and the tape speed. On a tape recorded at 30 frames (60 fields) per
second and at normal tape speed (15 in /s), each frame occupies 1/2 inch
of tape length, and, since the distance between the center of the erase -head
bracket and the headwheel is 7.5 inches, approximately 15 frames are
5 - ñó a 5:g
t?. -g cá n
l
17, E
03
CI. 00
$ Ñ
--
Ñ _á
Fig. 10 -11. Functional b ock diagram

"4'
ñ "g"
=f
O.
j
tré
s
H ó $ 1 ° â
.°S
Ñ i$
152,
0ô « o 0
Um Whò°
U
5 o
8
N
of RCA electronic splicing system.

Table 10 -1. Major Components of Electronic Splicing System

Major Circuits
Mounting
Component Location Designations Functions
Splice Timing Module Bank Electronic Automatically control starting and stopping
Module Counter, Erase times of record and erase functions during
and Record formation of ingoing or outgoing splices.
Delay Circuits
Final Binary Energize or de- energize output buses to

Switches other splice modules or auxiliary splice
circuits at times determined by counter and
delay circuits. Have two stable states, re-
cord and play.
Normalizer Keeps final binary switches in record state
Circuit when splice mode bus is off. This disables
splice circuits and places recorder in normal
condition.
Splice Logic Module Bank Switchlock Detects whether machine is in Switchlock
Module Detector during playback before splicing (splice play
mode). Prevents entry into splice record
mode if machine is not in Switchlock. If
machine is in Switchlock, informs operator
by lighting white SL indicator lamp.
Stop Memory Delays entry into stop mode until end of
Circuit and outgoing splice cycle, if STOP button is
Associated pressed during an insert splice recording.
Gates
Logic Circuits Prevent unlatching of record, play, wind,

or fast forward mode modules during out-
going splice cycle. Determine operation of
video -head transfer and servo-control cir-
cuits during splice operations.
Splice Control On Frame Splice Mode Allows machine to record normally even if
Module Behind Record Circuit splice mode switch is in a splice position,
Control Panel provided that RECORD button is pressed
before PLAY button. If PLAY button is
pressed first, circuit energizes splice mode
bus, which enables operation of splice cir-
cuits. Configuration of this circuit is similar
to mode modules in control system.
Play, Wind, Combines play, wind, fast forward buses
Fast Forward of control system into a single control bus
Matrix called the PWFFS bus. This bus starts out-
going splice cycle when PLAY, WIND, FAST
FORWARD, or STOP buttons are pressed
during an insert recording.
Two-Speed Modifies splice timing circuits for normal or
Switch half tape speed.
Reference Keeps reference generator, capstan oscilla-
Generator and tor, and guide servo in play mode when
Capstan splice mode bus is on.
Oscillator
Transistor
Switch
Video Head Perform additional logical operations on
Transistor output buses of splice logic module.
Switch,
C.T. and Cap
Error Switch,
Unlatch Inhibit
Switch
Table 10 -1. Major Components of Electronic Splicing System (Cont)

Major Circuits
Mounting
Selective Replace Selective erase head has separate sections

Master Erase Original for erasure of video tracks and control
Head and Master Erase track. Relay on bracket permits disconnect -
Erase Bracket Head and ing control -track section from source of
Bracket on Tape erase current. This permits retention of
Transport original control track for insert recording.
Auxiliary Audio Record Splice Audio The splice audio, cue, and video record
Circuit Module Record Switch switches permit inhibiting the record func-
Board SA101 tions when required during formation of a
splice even though master record, audio re-
Auxiliary Cue Record Splice Cue
Circuit Module Record Switch cord and cue record buses from control
system are energized. The switches are con-
Board SA105
trolled by output buses from the final bi-
Auxiliary FM Modulator Splice Video nary switches on the splice timing module.
Circuit Module Record Switch When the corresponding binary switch is in
Board SA207 the record state, the inhibition is removed
and the function is performed normally.
However, when the binary switch is in the
play state, the inhibition is applied. When
splice bus is off, normalizer circuit on
splice timing module places all binary
switches in record state. This removes all
inhibitions and allows recorder to function
normally.
Auxiliary On Mounting Cue Retain Permit retaining original audio or cue
Circuit Frame of Splice Transistor tracks while recording new video tracks.
Board SA14A Control Module, Switch Audio and cue sections controlled by sep-
Behind Record Audio Retain arate toggle switches which replace origi-
Control Panel Transistor nal RELEASE push buttons on record control
Panel. When audio or cue section is off,
Switch
master record bus is disconnected from
audio or cue record bus, and erasure and
recording on corresponding tracks is pre-
vented while machine is in master record.
Auxiliary Control Track Filter and Prevent abrupt change in control -track level
Circuit Record /Play Multivibrator when machine is switched from play to
Board SA319 Module Clamp Circuits record.
Auxiliary Capstan Error Capstan During add -on splice recording, matches
Circuit Module Memory Circuit tape speed to that of original recording.
Board SA321 Not used during insert splice recording be-
cause tape speed is referenced to original
control track.
AUDIO and Toggle Toggle switches, in conjunction with auxil-
CUE RECORD Switches Re- iary board SA14A, permit choice of retain -
Toggle place Original ing original audio or cue tracks while re-
Switches and AUDIO and cording new video tracks. Modified AUDIO
Indicators CUE RELEASE and CUE RECORD indicators provide both
Buttons on off and on indications.
Record
Control Panel
Erase Replaces Has circuit modifications which improve
Oscillator Original erase and bias waveform, reduce clicks due
Module Module in to mode switching, and permit audio spot
Module Bank erasure. Provides dc noise balance control
for cue channel, and auxiliary splice audio,
cue, and video erase transistor switches.
Erase switches inhibit erase functions when
corresponding final switches on splice tim-
ing module are in the play state. Inhibi-
tions are removed when final switches are
in the record state.
(Table continued on page 306.)

Table 10 -1. Major Components of Electronic Splicing System (Cont)
Major Circuits
Mounting
Indicators Mode Indicator SL Indicator Original SLindicator modified to include
Bank Above both red and white lamps. Red lamp lights
Tape Transport when machine is in splice play mode and
servo selector switch is on Switchlock.
White lamp lights when machine actually
enters Switchlock.
ADD Two sections each containing both red and
ON /INSERT white lamps. Red ADD ON lamp lights
when splice mode switch is on Add On.
White ADD ON lamp then lights when ma-
chine actually enters splice mode (splice
bus energized). Insert section operates in
similar manner when switch is on Insert.
SPOT ERASE Lights when audio spot erase circuit is
energized.
present. Actually, the master erase head is offset by about IV8 inch, so
slightly less than 15 frames are present between the heads. The offset is
provided because of mounting tolerances and circuit requirements.
Because of the preceding considerations, when the machine is set for 60-
field standards and normal tape speed, the counter is automatically set to
permit counting a maximum of 16 frame pulses just after the MASTER
RECORD button is pressed. Then, when the first frame pulse occurs, the track
on which video recording will start is exactly 15 frames in advance of the
headwheel.
At this time, the erase delay circuit is actuated. This circuit allows
enough unerased tape to go by after the end of the fifteenth frame has
passed the master erase head to create the desired overlap between the old
and new recordings ( normally adjusted for 11/2 tracks) It then causes the
.
video erase switch in the erase oscillator module to allow application of

current to the erase head. Since the splice mode selector switch is in the
add -on position, the control -track erase relay remains de- energized. There-
fore, the erase current is applied to both the video and control -track
sections.
During the remainder of the cycle, the capstan servo remains in the play
condition, and the tape speed is controlled by the unerased portion of the
original control track. No further action occurs until the fifteenth frame
pulse. At that time, the counter actuates an advance transfer circuit which
energizes the video -head changeover relay. The advance transfer allows time
for the changeover to take place before the video record current is turned
on. In addition, the advance transfer circuit cuts out the guide servo to
prevent a change in head -to-tape pressure during the recording.
At the sixteenth frame pulse (end of fifteenth frame) , the counter
actuates the record delay circuit and the splice -control final switch. The
record delay circuit delays video recording until the headwheel nears the
end of the track. The splice -control switch shuts off the counter and switches
the control -track record /play module to the record condition.
At the end of the record delay ( approximately 560 microseconds ) , the
splice video record switch on the fm modulator module allows the record
current to enter the video heads. At the same time, the audio and cue splice
switches go off and allow erasure and recording on the corresponding tracks
( unless the audio or cue toggle switches on the record control panel are in
the off position) .
Recording then proceeds normally except that the guide servo is cut
out, the capstan oscillator remains in the play condition, and the oscillator
is referenced to the output of the capstan memory circuit. ( Also, if the
CAP ADJ /FIXED TW /VAR TW switch is on Var TW, the headwheel phase
is shifted by the amount determined by the SPLICE RECORD HOR PHASE
control.) This condition of the machine is called the add -on splice record
mode.
At the end of the add -on recording, the operator may immediately switch
the machine to a desired mode by pressing the PLAY, WIND, FAST FORWARD,
or STOP button. If play is selected, the splice mode circuit remains on, and
the machine returns to the splice play mode; otherwise, the splice mode
circuit goes off, and normal operation is restored.
Insert Operation
The operating procedure for insert operation is the same as for add on,
except that the mode selector switch is placed on Insert. Circuit action dur-
ing the initial phases is also the same, except for the following:
1. When the PLAY button is pressed, the control -track erase relay be-
comes energized and disconnects the control -track section of the selec-
tive erase head. As a result, the control track will not be erased during
the recording.
2. At the end of the ingoing splice cycle, the recorder enters a condition
called the insert splice record mode. In this mode, the entire capstan
servo remains in the play condition, and the tape speed is governed
by the original control track. Also, unlatching of the master -record
mode module is prevented by an inhibit voltage, and the unlatch bus
is disconnected from the play, wind, and fast -forward mode modules.
An insert splice recording may be terminated by pressing the PLAY,

WIND, FAST FORWARD, or STOP button. This action starts an outgoing
splice cycle during which the master -record mode module remains latched
but the various recording and erase functions are shut off in the proper
sequence. In addition, the mode module corresponding to the button
pressed becomes energized, but the functions normally performed by this
module are delayed. (This delay is caused by a number of changes in the
control -circuit connections made during installation. One change consists
of reconnecting the wind relays so that they cannot be energized until the
capstan -roller solenoid is de- energized. Another consists of shorting the
sections of the STOP buttons in the stop ground interlock system to prevent
the mode modules from being turned off as soon as the STOP buttons
are pressed.)
No mode module for the stop mode is included in the original recorder,
but a stop memory circuit on the splice logic module serves the same pur-
pose. This circuit "remembers" that the STOP button has been pressed but
does not allow the machine to enter the stop mode until the end of the
outgoing splice cycle.
The actions occurring during an outgoing splice cycle, in 60 -field
standards and at normal tape speed, may be summarized as follows:
1. On the first frame -pulse count, the erase delay circuit is actuated. At
this time, the track at which video recording will stop is 15 frames
from the headwheel, and the point at which video erasure will stop
is 21/2 tracks closer to the headwheel (assuming that the ERASE
DELAY control was adjusted to produce an overlap of 11/2 tracks for
an ingoing splice.)
2. At the end of the erase delay period, the correct point reaches the
master erase head, and the video erase switch shuts off the erase
current. (The outgoing erase delay is automatically reduced by 4
milliseconds from the ingoing erase delay to produce an asymmetrical
overlap.)
3. At the sixteenth frame pulse (end of the fifteenth frame) , the
counter actuates the record delay circuit. After a delay of 560 micro-
seconds, this circuit actuates the video record switch and the cue and
audio erase and record switches, thereby shutting off the currents to
the corresponding heads.
4. The counter then starts another complete cycle of sixteen counts.
During this cycle, the master -record mode circuit remains on, but the
machine neither plays nor records.
5. At the fifteenth count of the second cycle, the advance transfer cir-
cuit de- energizes the video -head changeover relay and places the
guide servo in operation.
6. At the sixteenth count, the splice -control final switch stops the
counter, unlatches the master -record flip -flop, and allows the machine
to enter the selected mode. If play was selected, the machine returns
to the splice play mode. However, if wind, fast forward, or stop was
selected, the capstan -roller bus unlatches the splice mode circuit. The
normalizer circuit then restores the machine to normal operation.
NOTE: It should have been observed by the careful reader that the Ampex
and RCA systems have a different number of frames between the selective
erase head and the video head. This has nothing to do with interchange-
ability of tapes made on the two systems. It simply involves different

splicing logic for the individual electronic splicing systems.
10 -4. PRINCIPLES OF PROGRAMMED ELECTRONIC EDITING
We have covered the progress of splicing techniques from the razor

blade to the electronic method. Programmed electronic editing permits
a computerized, automatic programming type of operation.
Early (manually controlled) electronic editing removed the necessity
for cutting the tape and eliminated the audio /video offset problem. Using
devices to count edit pulses provided a means for identifying individual
frames. The pulses also provided proper timing for performing push-
button electronic edits. This process still left much to be desired, however,
and was primarily limited to "assemble" editing functions. The final break-
through to modern electronic editing came in approximately 1966 with
the addition of a serial time code. This code, recorded on the audio cue
track, provided a convenient means of frame identification at any point
on a given tape.
Modern electronic ( code controlled) editing has added several new
dimensions to the business of producing material for television. Easy edit -
point selection is facilitated through high -speed search of the recorded
material. Cueing to a selected edit point is a simple automated function.
Tape locations are easily logged, and, with frame -by -frame control, highly
sophisticated editing procedures may be employed. Complete preview be-
fore edit is easily implemented. Most important, these new systems increase
efficiency and reduce costs.
The remainder of Section 10 -4 consists of excerpts from "Creative VTR
Editing," by O. F. Wick and J. H. Frishette (RCA Broadcast News, Vol.
146, June 1971). These excerpts will serve to outline briefly how pro-
grammed electronic editing is used at NBC in Burbank, California. They
appear here through the courtesy of RCA.
In the spring of 1958, a new recording facility was installed at the NBC
Burbank studios. A total of 12 colorized video tape recorders were included.
With this expanded capability, we were in a position to undertake pre -
recordings that had not previously been possible. Within a sort time, the
need to edit and rearrange this material became very clear, and the possibili-
ties for doing so were investigated. The only implements available for the
cutting and splicing of video tape were rather crude: a milled channel, a
straightedge, and a razor blade. The splices produced invariably resulted
in severe picture disturbance-an intolerable situation if we were to
compete with motion pictures.
To improve our ability to assemble prerecorded tapes with acceptably
smooth transitions and to cope with other problems that had become ap-
parent, such as the handling of picture and sound offset, a study and
development project was initiated. This effort was quite productive, and
a procedure for the double- systemediting of video tape= ( independent

editing of pictures and sound) was evolved that has been in use to the
present time. Many of NBC's top shows have been assembled using the
tools and techniques that were developed in the course of this original
project. The Bob Hope shows and the Rowan and Martin "Laugh -In"
series are specific examples.
With the advent of the direct color recovery systems, high -band record-
ing, and numerous other improvements, the duplication of video tapes
became practical. This, in turn, made electronic editing feasible, but the
original equipment designed for this activity had a rather limited range of
usefulness -especially evident when complex sound -track manipulation
was required. Despite the limitations, however, a considerable number of
major productions have been assembled using this "first- generation" equip-
ment, both by us and by others.
Edit Control System
In 1967, a system of edit control was introduced that utilized a recorded
time code. The system not only provided a means for predetermining edit
points, but also included high -speed search and intermachine synchronizing
capability.
At about this time, we in Burbank were confronted with the need to
expand our editing facilities, so a careful study of the new system was
made. After considering as many of the factors as possible, we concluded
that the time -code system of control had many immediate advantages and
that there were potentially a number of other areas into which its useful-
ness might be extended. We decided to purchase two sets of this equipment
and to construct a special edit room to contain the two vtr's so equipped.
This installation was completed during the summer of 1968 and was used
on several shows during the fall season of that year.
After several months of experience -and a number of unexpected prob-
lems-the system was once again carefully examined, and once again we
concluded that it was the correct approach to the task at hand. Two addi-
tional sets were ordered and another special editing room constructed.
However, our experience had shown that with improved editing equip-
ment, more complete video and audio switching facilities immediately
available to the editors were also needed. So, the new room was designed
to include these. Also, the slow- motion disc had become an indispensable
accessory, and this, too, was provided in the room.
RCA Time -Code Editor

As the use of time -code editing systems increased, more effort on the
part of several manufacturers was devoted to the study of video -tape editing
2See Oscar F. Wick, "Double-System Recording and Editing With Video

Tape," Journal of the SMPTE, March 1960, p. 164.
problems. The advantages and disadvantages of the systems in use in our

plant were discussed in detail with engineers from RCA Camden. Their
investigations resulted in the design of a new system incorporating many
improvements in "human engineering" and providing automation for
several of the routine operations that must be performed repetitively.
Examples are: the entry of edit -point time -code data directly into memory
by depressing a single button, the performance of arithmetic calculations
within the editing circuitry, and precise recueing by depressing a single
button.
In view of the heavy schedule projected for the fall of 1970, still another
editing facility for NBC Burbank came under consideration early in the
year. It was decided that a third room complete with six -channel audio and
video switching equipment, video effects unit, slow -motion disc, and exten-
sive video and audio monitoring facilities would be constructed. It would
utilize the newly designed RCA time -code editing system (Fig. 10 -12) . In
September, this editing complex was placed in operation and assigned to
handle the postproduction requirements on the Dean Martin and Red
Skelton shows. The Dean Martin Show provides some interesting examples
of the capability of this equipment.
To begin, we should mention that it is standard procedure in our plant
to feed the time code to all machines at all times. Whenever a recording is
made, the cue track contains a time code representing the accurate time of
day. Normally, on a show such as the Dean Martin Show, the editor also
operates some of the recording equipment. This allows him to become
familiar with all of the details and to log all "takes" as they occur using
clock time. Later, during the editing session, these logged times are used
to locate any desired segment quickly, utilizing the high -speed search
capability of the system.
This high -speed search capability is also used in nonediting functions.
If we wish, for example, to play several segments of a football game in close
sequence into a live sports program, the unit programmer is set to cue the
game tape precisely before each segment. All the video -tape operator needs
TR -70/60 TR -70/60
VTR VTR
Auxiliary Unit Time -Code Unit

Programmer Programmer Generator Programmer
Auxiliary Auxiliary Auxi iary

Display Display Display
Fig. 10 -12. Typical two -machine time -code editing system.

to do is to enter the desired time into the unit programmer stop register
(by using the 10 -key keyboard) and press the SEARCH button. Generally,
a log of noteworthy activity is kept during the progress of these games,
and the use of the time code permits desired action to be cued reliably
without prior rehearsal.
Using the Time -Code Editor
On the Dean Martin Show, it is customary to record two separate feeds
from the originating studio. One feed carries the entire studio output, while
the second represents the unswitched output of the single camera normally
used for close -ups. Here again, the common time code supplied to both
sets of recorders permits the two pictures to be "intercut" during editing
while precise sound synchronization is maintained.
The principal elements of the Martin show are recorded on Saturday
evenings, and editing is scheduled for Monday and Tuesday. A total of
about 24 hours is required to complete the editing of the average show.
Usually, the editor and an assistant are employed in the editing process.
The editor operates the vtr on which the show is being assembled, as well
as the switching and effects equipment. The assistant handles the playback
vtr, its programmer, and the slow-motion disc equipment. When dissolves
between reels are used, a second assistant and a third vtr are assigned.
A 64-minute reel containing previously recorded control- track, black-burst,
and time -code ( from 0 to 60 minutes) signals is used as the assembly
medium. This becomes our electronically edited master, and all edits are
in real show time.
When "freeze" is employed, picture and sound from the segment to be
frozen are assembled in the normal manner on the master reel. While this
is being done, the picture is simultaneously recorded on the disc. The pic-
ture on the master reel is then replaced by a replay from the disc, with
action stopped at the appropriate point. It is, of course, essential that pre-
cise synchronism be maintained between the vtr and disc, since the disc
picture must exactly match the previously recorded sound up to the "freeze"
point. To achieve this necessary condition, use is made of a contact closure
that is externally available from the unit programmer. This closure is used
to start the disc from the master -reel time code when the material is
recorded, and, consequently, an exact time relationship between vtr and
disc is insured for playback.
Time -code -actuated contact closures are used to control a variety of ex-
ternal equipment in addition to the slow-motion disc. Sound playback
machines, auxiliary vtr's, and automatic effects units are frequently started
and switched by this means.
A and B Rolls
On the Martin show, opening and closing titles are produced during the
editing operation. Separate A and B rolls are prepared by careful assembly
from selected portions of the original recordings, and these are played back
in synchronism utilizing the synchronizing feature of the unit programmer.
The standard closing for the Martin show is 60 seconds long and consists
of roll titles matted over a series of stills followed by a segment in normal
motion. Finally, the scene fades to black. Sound during the closing is the
show theme. In a first step, as indicated previously, the editors prepare
A and B rolls of the entire series. These are rolls of tape that contain suc-
cessive show segments so arranged that, in playback, dissolves may be made
from roll A to roll B, and vice versa, and thus produce the sequence
desired. The stills required are obtained by freezing selected frames from
the original recordings. In making the assembly, a third vtr, also started
by contact closure in the unit programmer, is used to play back previously
recorded roll titles. All picture signals are routed through the switching
equipment in the room so that the roll titles may be matted and the dis-
solves executed. Theme music is also "laid down" at this time. The time -
code editing system has substantially reduced the time and effort required
to produce a sequence of this type.
Sweetening the Sound

When shows are edited, it is generally necessary to reprocess the sound
track because small but objectionable changes in background level and
quality often occur at the edit points. This is especially noticeable when the
splice is made during applause or laughter. Then, too, it is sometimes neces-
sary to add sound effects, musical bridges, and audience reaction (if an
audience was not present during the recording) . The Dean Martin Show
is performed before a live audience and the response is quite adequate.
However, "sweetening," as it is called, is necessary to obtain a complete,
smooth, and uniform finished track.
The average Martin Show contains 15 to 20 sketches, songs, etc., and
each of these frequently requires some internal editing, making for an
average of about 90 splices in the completed master tape. During sweet-
ening, the jumps in the edited track are bridged by sound similar to that
already existing. For example, if the edit occurs in applause, similar ap-
plause would be mixed in to cover the discontinuity.
After editing of the show has been completed, the sound track and time
code are transferred to two tracks of a four -track audio recorder. This
recorder is of the type employing a capstan servo and sync track, so syn-
chronism with the vtr is absolute. During the sweetening session, the four -
track audio machine is used in the playback mode to provide program
sound and is also used as the source of the time code to which the vtr is
synchronized by the unit -programmer synchronizing feature. This results
in a situation wherein the picture is supplied from the edited master tape,
and the program sound is supplied from the four-track recorder, both in
precise interlock. These signals are routed to a postproduction studio con-
taining extensive audio mixing and equalizing facilities as well as excellent
audio and video monitoring. Any necessary modifications to the original

track are made in this studio, and the composite sound output is fed back
to and recorded on the vtr as the final sound track.
10 -5. THE EECO EDITING SYSTEM

Although the specific Electronic Engineering Company equipment de-
scribed in this section includes refinements over some simpler systems, a
brief description of it will serve to illustrate fundamentals of all modern
pogrammed editing systems. This equipment is designed specifically for
use with RCA TR -60/70 and Ampex VR- 1200/2000 series television re-
cording systems. The basic equipment consists of an electronic editor
programmer and a time -code generator. Accessory equipment includes an
auxiliary programmer, an auxiliary time display, and a remote transport
control panel.
The Edit Code
An edit code is the key to efficient indexing and electronic editing of
video tapes. Essentially it is a binary time code that is recorded on the
cue channel or the second audio channel. The SMPTE edit code uses-80 bits
per frame to carry the following data:
1. Hour, minute, second, and frame count
2. Eight optional four -digit words for additional information or control
3. A sync word to indicate the end of each frame and the direction of
the tape
Edit -code time is based on the frequency of line power or the video /sync
input. The SMPTE edit code (Fig. 10 -13) offers control capability for the
fully automated editing of the future.
Color (NTSC) video requires a "drop- frame" technique not required
with monochrome recordings. The actual NTSC frame rate is 29.97002618
per second, hardly as practical a time base as 25 or 30 frames per second.
In working with NTSC video, the edit -code time reference is set at 29.97 Hz
(or frames per second). To prevent a resulting time -error accumulation
of 86 seconds per day, two frames are dropped every minute except on the
tenth minutes. The drop -frame reference signal is shown in Fig. 10 -13.
The edit code contains the generated time of day in binary-coded decimal
(bcd) form, the frame count in bcd form, a sixteen -bit sync word, and
spaces for eight optional four -bit binary words. The time of day in hours,
minutes, and seconds is represented in six code segments of four bits each.
The frame count is represented in two code segments of four bits each.
A logic 1 in the eleventh bit of the code indicates a code that drops frames
for correction. The sync word is a unique bit pattern that cannot be dupli-
cated by other information in the time code. It is used to indicate the end
of each code frame and to show the direction of tape motion. The eight
if - v É
_
1(
Fig. 10 -13. Example of SMPTE edit code for one frame.

optional binary words may be used to convey additional information when

the binary word option is included.
Each 80 -bit code frame has a duration established by the video frame
rate. The code frame is updated and repeated once for each video frame.
A transition from low to high or high to low occurs at the beginning of
each bit. The code is therefore self -clocking. A binary 0, or false, may be
either high or low. A binary 1, or true, is represented by a transition
occurring at the midpoint of the bit. This change of state from high to
low or low to high results in a pulse of 1/2-bit duration. (See upper right -
hand section of Fig. 10 -13.)
Tape Indexing- Editing "Mini -Modules"
The EECO "Mini-Module" units are normally used in pairs. Fig. 10 -14
shows two of these units. The BE 520 (not illustrated) generates an elec-
tronic signal (edit code) that is recorded on video tape. Either the BE 420
or BE 400 (right and left, respectively, in Fig. 10 -14) can read this signal
as taped video scenes are monitored, and either unit provides a visual dis-
play of the edit -code signal expressed in terms of real or elapsed hours,
Courtesy Electronic Engineering Co. of California

Fig. 10 -14. EECO video character generator (left) and edit-code reader.
minutes, seconds, and frames. All three units are small (approximately 5"
X 8" X 11 ") , hence the term "Mini-Modules." Basic applications are out-
lined by Fig. 10 -15. The units are completely solid state, using integrated
circuits and transistors.
The BE 520 Edit -Code Generator

The BE 520 generates the indexing code needed for efficient editing of
video tapes. The code is in the form of an electronic signal recorded on the
cue channel or second audio channel and represents real or elapsed time
expressed in hours, minutes, seconds, and frames. Normally the edit code
is recorded on the tape as scenes are shot, but it can also be added to pre-
viously recorded tapes.
(A) Recording (code generator) .
Edit Code Video

BE 420 BE 400
Edit -Code Video Character
Video Edit Code
Reader Generator
Display
Monitor
14233821
Display
(B) Reviewing programming Reviewing programming

(code reader) . (character generator).
Fig. 10 -15. Basic applications of indexing -editing equipment.
Some of the advantages and options available when video or audio tapes
carry the edit -code signals are:
1. Tapes can be previewed and a sqeuence log written to show start and
stop times of selected scenes.
2. Subsequent program editing, either manual or computer- automated,
can be made more quickly and at less cost than with other methods.
3. Scenes to be shown in a live program can be located more rapidly
either manually or automatically.
4. The edit code can be used to synchronize dual-equipment operations.
5. Recordings of significant events carry a permanent time record.
Sequences and single frames can be examined in terms of actual
time and with split - second accuracy.
6. The edit code can accompany the transfer of video scenes from one
type of recorder to another. For example, quadruplex recordings
can be copied by a helical recorder. The low -cost equipment can then
be used for careful scene reviewing and development of a program
sequence log. Subsequent editing of the original quadruplex tapes
can be made with minimum use of production -studio time, including

automatic or computer -automated editing equipment.
7. Elaborate computer storage, retrieval, and editing systems are already
used for some types of video programs. The edit code is used for
reference indexing, storage retrieval, and program sequencing.
8. The edit -code index is never dependent on the original tape length
and mechanical footage count. Any section of tape carries its original
edit -code identification.
The generator utilizes either the power -line frequency or an externally

applied video /sync signal as its reference frequency when generating time
at frame rates of 25 or 30 Hz, and the time of day coded into the output
time code is then as accurate as the selected reference frequency. At the
29.97 -Hz color frame rate, however, the time of day will be in error. This
error is due to the fact that the generator counts as if the video frame rate
.
were 30 frames per second, or 1800 frames per minute. However, at the
rate of 29.97 frames per second, only 1798.2 frames per minute are gen-
erated. Therefore, an error of 1.8 frames per minute (1800 -
1798.2 = 1.8)
+60
40
.a.,,I.im..,
IFAMIA I
+20
20
40
_60
0 2 3 4 6 7 8 9 10 11 12 13 14 15
20 30
40 50
Time (Minutes)
Fig. 10 -16. Time -error correction.
is induced. Because the basic video frame duration is 1/30 second, or ap-
proximately 33.33 milliseconds, the time error is 60 milliseconds (1.8 x
33.33 = 60) per minute. To overcome this error, the generator drops two
frame counts per minute for a time correction of 66.66 milliseconds (2 x
33.33). However, this effectively overcorrects the error by -6.66 milli-
-
seconds (60 66.66 = -6.66) . (See Fig. 10 -16.) At the end of the
second minute, the total error before correction is approximately 53 milli-
seconds. After the two -frame correction, the total error is approximately
-13 milliseconds. The two -frame correction is repeated until after the ninth
minute correction, when the total error is approximately -60 milliseconds.
Between the ninth and tenth minutes, the 60- millisecond error generated
7
- Ua
d
à
d
°0
U
2p .J
m
C or \
T 1
tot a,
L t0
CD
IÓ O
a3 u
E
N O
o
j
m
CL
-O
U ..°..
L
H
°w
CC
0 u
` T CO
LL N G
N
S
O
.O
_O
Fig. 10 -17. Simplified block diagram of EECO BE 520.

returns the total time error back to 0 milliseconds. Therefore, it is unneces-

sary to make the two -frame correction at the tenth minute.
Since the color video frame rate is actually 29.97002618 instead of
29.97, there is actually an accumulated error of 0.51 millisecond in ten
minutes. This amounts to a total error of 75 milliseconds (0.075 second)
per day. If the two -frame error correction were not made each minute ( ex-
cept the tenth) , there would be an error of 86 seconds per day.
The generator utilizes either the power -line frequency or an externally
applied video /sync signal as its reference frequency. The video /sync signal
is applied through either connector J1 or J2 (Fig. 10 -17) to the frame sync
separator, which separates the sync information from the incoming signal.
The internal power supply operates from either 117 or 234 volts, at 50 to
60 Hz, to provide the necessary dc voltages required by the circuits in the
generator. The power supply also generates a 50- or 60 -Hz signal (depend-
ing on the operating power-line frequency) , which is applied to a squaring
amplifier. The squared signal is applied to a flip -flop which generates a
signal at half the line frequency. The source- switch circuitry selects either
the video /sync or 1/2- power- frequency signal and applies the selected signal
to the phase comparator as the reference -source signal.
The phase comparator compares the reference -source signal with an
output signal from the phase comparator pulse circuit. The frequency
difference, if any, is converted into an error voltage which is applied to the
voltage controlled oscillator (vco) The vco is a free -running oscillator
.
the basic frequency of which is changed by the error voltage. A derivative

of the vco frequency is fed back to the phase comparator. This forms a
phase -lock loop which synchronizes the vco output frequency to the selected
reference -source signal. A no -sync indicator circuit gives a visual indication
when the vco is not synchronized.
The output frequency of the vco is divided by the code-bit counter,
producing various code -bit signals and frame -rate signals. The code -bit
signals and clock signals are applied to the scanners to generate the sync-
word and frame information for the edit code. A one -pulse -per- frame -rate
(1 ppfr) signal is applied to the frame counter to update the frame count.
An inverted 64- code -bit (64CB) signal is also coupled through an output
buffer to J4 as a 1 ppfr output signal.
The frame counter generates frame -count information. When the
reference frequency is a 30 -Hz video /sync signal, a 29.97 -Hz color video/
sync signal, or a 60 -Hz power -line frequency, the selected frame count is
30. When the reference is a 25 -Hz video /sync signal or a 50 -Hz power -
line frequency, the selected frame count is 25.
The outputs of the RUN, HOLD, and PRESET switch circuits are applied
to the accumulator control. The control signals reset all the accumulator
outputs to zero, preset the accumulator output at a 50 -kHz rate to any
desired preset time, advance the accumulator at a 1 -pps rate, or hold the
accumulator from counting.
The accumulator generates the time of day, 0 to 24 hours, at a 1- second

rate. This time is constantly available at the accumulator outputs in parallel
bcd (1- 2 -4 -8) form. The accumulator generated time is displayed by
numeric indicator tubes. Also, the generated time and the frame count are
applied through output buffers to connector J4 as parallel bcd signals.
The scanner receives the time -of -day information from the accumulator,
frame information from the frame counter, and signals from the code -bit
counter to produce the edit code. The scanner also generates a sync word
which is included in the edit code. The edit code is applied to an edit -
code output circuit for the necessary amplification, buffering, and imped-
ance matching before being applied to output connector J3.
Space does not permit a detailed description of the circuitry in this
equipment. However, a brief analysis of the phase comparator and the vco
will assist the reader in becoming acquainted with the type of logic circuitry
involved. See Fig. 10 -18.
The voltage -controlled oscillator (vco) generates the basic frequency
used throughout the generator to form the code bits and clock pulses. The
vco is a free -running relaxation oscillator the frequency of which is con-
trollable by the input dc voltage level from the phase comparator. The vco
output frequency is divided by circuits in the code -bit counter and then
reapplied to the phase comparator at a one- pulse - per -frame rate. Therefore,
a phase -lock loop is formed. When the phase -comparison pulses are com-
pared with a one -pulse -per- frame -rate reference-source signal, it is possible
for the phase comparator to lock the vco frequency accurately to the in-
coming video /sync signal or power line.
The phase -comparison signal is a positive -going pulse having a duration
of 32 bits and a frequency of one pulse per frame (Fig. 10 -19) . The posi-
tive phase -comparison signal causes Q11 (Fig. 10 -18) to conduct at a
constant current determined by R47. This constant current causes the
potential on C13 to decrease at a linear rate. This forms a negative-going
linear ramp with a duration of approximately 400 microseconds. This
pulse is applied to the source (S) of FET Q8, through emitter follower Q7.
A narrow reference -source pulse, derived from either the video /sync
signal or the power line, is applied to the base of transistor Q12. When
the reference -source pulse goes low, the collector of Q12 goes high, thereby
back- biasing diode X6. With X6 back -biased, there is no current through
resistor R43. Therefore, the voltage difference between the source and the
gate of Q8 is zero, causing Q8 to conduct. Capacitor C15 then charges to
the voltage at the source (S) of Q8. Because Q8 is turned on at the frame
rate (25 -30 per second), the voltage level at C15 is maintained and applied
to the gate of FET Q9. This FET functions as a source follower, so the
voltage level established at its gate also appears at its source.
The oscillator portion of the circuit consists of unijunction transistor
Q10, capacitor C17, inductor L1, resistor R35, and potentiometer R37.
Capacitor C17 charges through R35 to a level that will allow Q10 to fire.
r
VCO
5V
R37
BI
Cl?
Phase Comparator
841
Q9
C IS
00 X6 012
TP3
cui
R36
07
1111
R47
+I2V
111 +5V
+5V.
Phase Ref
Ref Source
Phase Comp
+12V
No Sync
112 of IC8 1/2 f IC8

Q ,Retriggerable
One soot Q
400ps One Shot
50 rns
No Sync Indicator
Fig. 10-18. Simplified logic diagram showing

Code Bit Counter
CLK 1
K 0
IC37 p
CLK
CLK
O
1/2 CB
CL
OR
ICB
ICB
CLK
CR
2C
2CB
4C
IC38 BCB
(Same as Atom/ 16CB
32C B
32CB
Reset Latch
80CB
IPPFR
61(
-d CLK
MCB
phase comparator and vco section of BE 520.

Capacitor C17 then discharges through Q10 and L1. When the capacitor
is discharged sufficiently, Q10 turns off and C17 begins charging again.
Potentiometer R37 determines the firing point of Q10 and thereby adjusts
the frequency of the oscillator.
Capacitor C17 receives an additional charge through Q9. Therefore, the
point on the ramp at which Q8 fires also determines the frequency of the
vco. Because a derivative of the vco signal is utilized by the phase com-
parator, the oscillator is self- compensating and will lock onto the refer-
ence- source signal.
32
Phase Comp
Q8 (S)
Q8 (Gli
N
I
Test Point 3
-
(Expanded)
4GOys --i
Fig. 10 -19. Phase- comparator waveforms.
Test point 3 provides a convenient point for monitoring the adjustment

of the vco frequency. With an oscilloscope, both the reference - source
pulse and the phase -comparison- signal ramp can be monitored. Poten-
tiometer R37 is adjusted to position the reference-source pulse to coin-
cide with the ramp (Fig. 10 -19).
In addition to the actions described above, a phase- reference pulse is
coupled through diode X4 to the no -sync indicator circuit.
The BE 420 Edit Time Code Reader
The BE 420 "reads" the standard SMPTE edit code from any source
and displays it as real or elapsed hours, minutes, seconds, and frames.
This display is available as reference during tape play, rewind, fast for-
ward, or single -frame hold. Scenes from quadruplex or helical tapes may
be reviewed and a program sequence developed to show the start and
stop times of selected scenes. The program sequence provides an accurate
index for manual, automatic, or computer- automated editing of the tapes.
The BE 420 also has helpful supplemental outputs. For example, a

field -rate output can be used to keep a multitrack audio recorder in sync
with video playback; this eliminates the need for a separate control track
in double- system operations (Fig. 10 -20) . Parallel edit -code outputs are
used in computer /automated editing and other equipment controls.
A hold button allows displayed time to be frozen at any point to aid
scene identification. Once logged, any scene can be quickly relocated for
inclusion in a live broadcast. Significant events can be analyzed in terms
of real time.
Edit Code
Multitrack From Control Track BE C ?0
Audio Edit Code
Recorder Reader
Fig. 10 -20. Method of synchronizing Reference
audio recorder having control servo. Derived From Edit Code
Studio Reference
The incoming SMPTE edit time code is applied through connector J1

to the demodulator (Fig. 10 -21) . The demodulator utilizes the edit time
code to produce two signals, data ones and code clock. The data ones is a
series of positive pulses, each being I/2 bit count in duration and each
occurring when a one transition occurs in the incoming edit time code.
The code clock is a series of 80 positive pulses per frame, each with a
period of one code bit and a duration of approximately one microsecond.
Each code clock pulse is timed so that it occurs just prior to the end of
a data ones pulse (if present) .
The data ones and code clock signals are applied to a decoder. Both
signals are inverted (identified as CLK and ONES in Fig. 10 -21) and
applied to buffer amplifiers for output signals at J2. Two code -bit signals
(4CB and 64CB) are derived from the code clock and applied through
buffers to output connector 12. The sync -word portion of the data ones
signal is decoded (FRWD) and also applied as an output voltage level
at J2 through a buffer. Eight enabling signals (labelled 1F EN, 10F EN,
1S EN, lOS EN, 1M EN, 10M EN, 1H EN, and 10H EN in Fig. 10 -21)
are produced by the decoder and applied to counter /registers to distinguish
properly the various data ones (frames, seconds, minutes, or hours) from
each other. The decoder is capable of determining the tape direction (for-
ward or reverse) and, by utilizing a left -right shift register, transferring the
decoded serial data (1, 2, 4, 8, and enabling signals) to the counter/
register. In addition, field -rate, frame -rate, and read- enable signals are
produced, buffered, and made available at output connector J2.
The frames /seconds counter /register utilizes frame and second enabling
signals (1F EN, 10F EN, 1S EN, 10S EN) and the decoded serial data
(1, 2, 4, 8) to produce a set of parallel bcd outputs (labelled BCD
0
É
°'
c
á
-7-- " o` 'm ó v
3
i.° E
1i111ì1111
00_U
L)
03 co
V v
2O O
= C_
cc U
O
1S EN, IOS EN
OFF-ETI
Frame Rate
U o
U a
2
li
11
Fig. 10 -21. Block diagram of BE 420 code reader.

FRAMES and BCD SECONDS in Fig. 10 -21) , which are applied to the
frames and second display indicators and to buffers for outputs at J2.
Other signals used by the frames /seconds counter /register are (as labeled
in Fig. 10 -21) FWRD, FRAME RATE, and READ EN. A switch in the
:
frames /seconds counter /register selects the frame count (25 or 30) to
which the bcd frames are permitted to register before being reset. This
switch must be set by the operator to correspond to the number of frames
per second in the incoming edit time code.
Similarly, the minutes /hours counter /register produces parallel bcd sig-
nals (labelled BCD MINUTES and BCD HOURS) for application to the
minutes and hours display indicators and to buffers for outputs.
The HOLD switch activates a latch in the decoder which prevents any
further decoding of the ones in the decoder or changing of signals from
the counter /register. Therefore, in the hold mode, the display and output
signals represent the time and frame count at the instant the HOLD switch
was pressed. When the switch is pressed a second time, the latch is re-
leased, and the decoder, counter /registers, and display are permitted to
update to the time represented in the edit time code. The read -enable
signal darkens the frames display by disabling the high voltage to the
frames indicators.
The BE 400 Video Character Generator
The BE 400 video character generator converts any number of video
monitors into time -code displays. During video play or search, the unit
reads the edit code from the tape and generates a corresponding visual
time display on the monitor screens. Front -panel switches permit rapid
changes in the size and location of the time display, and a black back-
ground mask can be added or deleted.
Use of the unit is flexible. It can display the tape edit code (time) on
the monitors without affecting the video signals being broadcast or being
transferred to another tape. A common option, however, is to add the
visual time display to the video signal when it is transferred from a
quadruplex recorder to a helical recorder. Scene review is then easy be-
cause all scenes, moving or single frame, carry the visual time display.
The features just described make it possible to prepare, with relatively
inexpensive equipment, a program sequence log to show the start and
stop times of selected scenes. A start -stop time log is useful in a number
of operations:
1. Automatic electronic editing

2. Manual electronic editing
3. Computer /automated electronic editing
4. Rapid location of scenes for inclusion in a live program
5. General indexing and filing of taped materials
6. Indexing of computer -stored material
The BE 400 provides a simple method of putting a real -time display

on a live video broadcast. It can also provide simultaneous and accurate
time references on a number of remote monitors.
This unit accepts either a standard edit code or parallel time code and
frame number input. Control of the output accommodates a wide range
of viewing requirements. The BE 400 uses nonadditive video mixing to
avoid distortion in video processing channels. Inputs and outputs are
TTL, RTL, or DTL compatible.
The BE 210 Electronic Editor Programmer
The BE 210 is the heart of the EECO electronic editing system. It pro-
vides precise control of search, cue, and splicing functions, and is designed
specifically for use with the RCA TR-60/70 and Ampex VR- 1200/2000
series television recorders. This unit provides transport control based on
selection of predetermined start and stop points and contains the neces-
sary circuitry for programming the recorder edit electronics and recue
functions. This circuitry ensures that the edited scene will be the same as
the preview. Features are:
1. Electronic splicing function performed with equal ease for audio,
video, or audio and video combined
2. Separate preview and edit functions
3. Automatic frame-by-frame synchronization capability
4. Single sequence control to automate routine editing operation
5. Simple push- button touch entry for edit times and control
6. Keyboard entry and /or modification of start and stop times
7. Human -engineered control panel
8. Backlighted control panel for operation in subdued ambient lighting
9. Single glide recue and search operation
We will go through the basic functions of the BE 210 programmer for
the purpose of illustrating how modern electronic editing is computer-
ized. All buttons and switches mentioned are shown in the keyboard draw-
ing of Fig. 10 -22.
r
óö
rIX
°°111.,,..
mn
, 0a
O
El
6 H
,22,35,59,11,
MOM IMIY/FS WOOS
EDIT
s..wa.
L 7C C.JI
o O®
Courtesy Electronic Engineering Co. of California
Fig. 10 -22. Keyboard of BE 210 programmer.
Modification of Times-At any time, as required by the operator, the

start and stop times may be recalled and displayed. The start or stop time
can be modified by means of a simple front -panel keyboard assembly. The
keyboard is enabled by operating either the + or - button. This operation
automatically selects the frame number for modification as displayed on
the front panel. If modification of the hours, minutes, or seconds is re-
quired, the proper digit is selected by operating the left arrow button. The
proper digit is indicated by two periods located just below the digit.
Keyboard Time Entry -If times are to be entered from a log, the dis-
play register is cleared by operating the CLEAR button. This operation re-
sets the display to zero and places the locating periods below the tens -of-
hours digit. As the times are entered from the hours to the frames, the
indicator automatically advances to the next digit to the right. Operating
the NORM button returns the BE 210 to normal operation of displaying
the input code.
Sequence Control -In addition to the normally manual operating con-
trols, an edit sequence control is provided for normal routine edit opera-
tion. The normal edit sequence consists of play, start enter ( time) , stop
enter ( time) , recue, preview, and stop.
Each time the EDIT SEQUENCE button is pressed, the function indicated
by a lighted arrow will be initiated. The arrow beneath the next function
in the sequence then lights to indicate to the operator the next event in
the series. After the stop function has been initiated, the next sequence is
recue, preview, stop. This sequence repeats until the edit button is op-
erated. The edit will then be recorded as previewed. The sequence may
be interrupted at any point by simply pressing the appropriate function
selector.
Limited-Range Synchronizer-The purpose of the limited -range syn-
chronizer is to provide exact frame synchronization of the associated trans-
port to a reference time code. This capability can be used in transfer edit-
ing to synchronize the playback transport accurately to the record trans-
port by referencing to the time code from the record machine.
The limited -range synchronizer has an offset mode in which it will
lock the stop -time frame number from its own code to the start -time
frame number from the reference code. That is, the synchronizer will
phase the playback of its own code in such a way that when the code
frame coinciding with its start-time frame occurs, the reference code will
be at a code frame coinciding with the stop -time frame. The two codes
will then be locked in an offset fashion with the offset determined by the
respective stop and start times. This assures that the start of action of the
playback- machine tape matches the ingoing edit point on the record -
machine tape.
The limited -range synchronizer also has a nonoffset mode which dis-
regards the start- and stop-time frames and locks frame 1 of the con-
trolled code to frame 1 of the reference code.
Controls are also provided to advance or retard the capstan servo con-
trol circuit manually.
Additional front -panel controls provide the following functions:
PROGRAMMER ON: Activates control functions of the BE 210 to con-
trol associated video recorder.
PROGRAMMER REMOTE: Delegates control of BE 210 to remote point.
PLAYBACK /EDIT: Six mechanically interlocked buttons.
PLAYBACK CLOCK: The code at the recorder cue-track input is selected.
The time code generator (clock) is normally fed to this point. This
selection allows the transport (play and stop) to be controlled from
the time -code -generator time.
PLAYBACK SEARCH: The video -recorder cue -track output is selected.
The code at this point is one of two. During standby or record, the
code at the cue -track input appears at the cue track output through
the electronics -to- electronics (E -E) circuit. When the cue track is in
playback, the code is reproduced from the tape. This position is nor-
mal for tape search and cue. In this mode, the stop time is enabled
during recue. The tape may be searched and stopped or cued to any
point. A clock or slave code may then be used to place the recorder
into play (start) When the recorder is recued, the tape will be re-
.
wound to a point (10 seconds normally) ahead of the stop time.

PLAYBACK SLAVE: This external input is used primarily for transport
play control and is also the input from the reference machine for
the synchronizer. The second input to the synchronizer is from the
cue-track output of the local machine.
PROGRAMMER REMOTE overrides any selected playback input. When the
BE 210 is placed in remote, the code is selected remotely. When the clock
or slave inputs are selected and the recorder is placed into recue, rewind,
or fast forward, the code input automatically switches to the VR code in-
put. This is required because the playback code is needed for these opera-
tions.
EDIT A /V: Audio -video edit selection. Audio and video edits are made
simultaneously. The edit is begun at the start time (ingoing edit)
and terminated at the stop time (outgoing edit) In any of the three
.
edit functions, the edit sequencer is operative. When the recorder is

recued, the tape is rewound to a point (15 seconds normally) ahead
of the start time. A recue time longer than that of a playback ma-
chine is required to allow code to be read and a play command to
be given to the playback machine at the 10- second preroll time ahead
of the edit start time. The A/V function is used to start and stop
the preview or record of an audio -video edit. The code from the cue -
track playback is used in the edit functions.
EDIT A: Audio -only edits are made. See EDIT A/V above.
EDIT V: Video -only edits are made. See EDIT A/V above.
MANUAL: Operation of the manual edit button allows an ingoing edit
to be made using the edit button (after play) and an outgoing edit
to be made by operating either play, recue, preview, stop, rewind, or
fast forward. Operation of the manual button also resets the edit se-
quencer to the off condition (no arrows lighted, edit sequence but-
ton dim) .
START DISPLAY and STOP DISPLAY: Operation of the start (stop) dis-
play button transfers the time in the start (stop) storage register
into the display logic and displays this time. The time in the store is
not altered. This allows stored times to be read and modified. The
display is then restored to normal operation by one of three methods.
Operation of either the START ENTER, STOP ENTER, or NORM button
transfers the display back to the code input selector.
NOTE: The time in one storage register may be moved to another
storage register without destroying the time in the first. The NORM
button transfers the display to normal operation.
HOLD: The HOLD button freezes the display at the instant of opera-
tion. This feature is useful in displaying time without time entry.
The display is restored to normal operation with the NORM button.
The displayed time may be entered into the start or stop store if
required. Either entry button restores the display to normal.
Status Indicators- Indicators are provided on the upper left section of
the BE 210.
SYNC: Lights showing that the synchronizer has synchronized local ma-
chine to remote code.
No EDIT: Indicates edit was inhibited.
PREVIEW: Indicates that the video and /or audio monitors have switched
to the incoming program.
RECORD: Indicates that the video and /or audio monitors have switched
to the incoming program and the machine is recording.
START READY: Indicates that the start time function is enabled.
NOTE: Further editing and animation equipment involving video
disc recorders is covered in Chapter 14.
EXERCISES
Q10 -1. Why is it not possible to record a new scene into an existing program
simply by pressing the record button at the desired time while playing
a tape already bearing recorded material?
Q10 -2. Describe the basic difference between (A) the insert mode of splicing
and (B) the add -on (assemble) mode.
Q10 -3. What causes a momentary horizontal "lurch" or "bobble" in the pic-
ture at a splicing point?
Q10 -4. What controls the capstan servo in the recording machine (machine
on which editing is being accomplished) during (A) the insert mode
and (B) the add -on (assemble) mode?
Q10 -5. Basically, what is "programmed electronic editing "?
CHAPTER 11
Quadruplex Tape System

Operations
Television tape systems, now refined to a rather high degree of per-

formance, utilize many unique mechanical and electronic arrangements,
each of which makes an important contribution to the final production.
Any complex system of this nature requires the development of a certain
amount of skill in setup and operation.
Obviously, the degree of skill and knowledge required for operation
depends on the operating practices of a particular station. If the tape op-
erator is charged only with threading, selecting the proper video and audio
feed, and starting and stopping the system on cue, he will likely not be
interested in this book, except out of curiosity or as a means of advance-
ment. However, there are two more common operating practices, which
may be listed as follows: (1) In one type of operation, tv tape operators
are charged with the responsibility of cleaning, of adjusting modulator fre-
quency and deviation, of calibrating and adjusting video and audio levels,
of performing head optimization, of degaussing transport and heads, and
of checking out the system with a short trial recording and playback. In
this type of operation, maintenance personnel are called to service the
equipment in case of abnormal operation and may also be required to
carry out more time -consuming functions, such as head optimization, in
older tape systems. (2) In the second type of operation, the tv tape oper-
ators are charged with all the responsibilities under 1, plus all preventive
and emergency system maintenance.
All of the normal operational functions listed under 1 will be consid-
ered in this chapter. Chapter 12 will cover preventive and emergency
maintenance as well as system performance measurements.
11 -1. TERMINOLOGY
The terms most commonly used in tape production techniques may be
defined as follows:
332
QUADRUPLEX TAPE SYSTEM OPERATIONS 333
Original: The first tape recording of a given program. If multiple ma-

chines are used to record this signal simultaneously, then a number of
originals exist.
Master: The first tape recording that is complete in all production ele-
ments. An original recording which is complete in all production elements
can also be the master. A master may also be either of the following:
1. A tape which consists of any number of different originals (or
copies) spliced together.
2. A tape which is recorded from any number of other tape- system play-
backs (and perhaps from other sources such as film and live inserts)
fed through a switching system just as any original production is
performed. In this case, all tape systems and any other sources used
must be under the control of a common sync generator such as the
Ampex Intersync or the RCA Pixlock. This technique eliminates
the necessity of mechanical or electronic splicing.
Copy: A recording made from another tape signal, whether a master or
a copy. The term copying master is sometimes employed for a master tape
which has been designated for use in making copies. A first generation
copy is a copy made from a master. A second generation tape is a copy
of a copy, etc.
Leader: The portion of a tape immediately preceding program content,
usually containing visual and /or aural cues. The term is also applied to a
tough protective strip before the leading edge of the recording tape, used
to protect the first part of the tape.
Run -Out: The portion of tape following program content, usually con-
taining only sync information to allow fade or cut -out before going to
"noise" ( tape playback without signal) .
Standby Tape: Production term given by the director to cue the oper-
ator to get ready to start the tape rolling. In many instances, the operator
starts the capstan action (capstan turns without the tape being engaged)
on receiving this cue. This is always done when the fast -start method of
operation is followed. This method allows only 2 or 3 seconds after the
tape is rolling to the actual "take" by the video switcher. The tape- recorder
switch for this function is labeled OVERRIDE by Ampex and STANDBY by
RCA. This procedure is also sometimes practiced even when the more
normal 6- second lapse between roll and take is used, since the system
normally stabilizes more quickly when the capstan is already up to speed
before engaging the tape.
Roll Tape: Term used by the director to instruct the operator to start
the tape rolling by depressing the PLAY button. The capstan engages the
tape immediately, and a video signal is delivered from the tape as soon
as the vacuum guide engages the tape with the rotating heads. This inter-
val is normally 11/2 to 2 seconds, depending on the setting of the asso-
ciated time -delay relay. (This time may be as long as 4 seconds.)
Cue -In: Method used to enable the operator to stop the tape at the
proper point prior to the start of actual program content. Visual or aural
cues may be used; the aural portion is either on the regular program
sound track or on the cue track. One common visual method is the tele-
vising of a second -clock with the second hand started about 15 seconds
prior to the program, turning counterclockwise toward zero. The signal
is brought to black level 2 seconds before program start to provide a
clean take interval for the video switcher.
11 -2. THE SYSTEM CONCEPT IN TAPE OPERATIONS
Tape- recording setup and operational techniques become more diffi-

cult as the limits of use for which the tape is made are expanded. A pro-
gram to be recorded one day and run the next day on the same head may
be considered a highly limited application of responsibilities. Recorded
characteristics may be entirely incompatible with other heads or recording
systems but result in entirely satisfactory playback in this instance.
To go a step further, consider the case where a number of spot com-
mercials are to be recorded for use at the same station. Even though the
same tape systems will be used for these spots, there is certainly no guaran-
tee that the same head will be used throughout the life of the commercial
commitment. Some degree of standardization is now highly desirable.
The next step is interchangeability with other heads and other systems.
Important standardizations may be listed as follows:
1. Video Heads: Quadrature and gap- azimuth alignment, vacuum -guide

position horizontally (controlling tip penetration) , and vacuum -
guide position vertically (for correct concentricity at the standard
tip penetration).
2. Video Signal: Carrier frequency, devil- Lion related to 100 -percent

modulation, type of pre- emphasis, and video bandwidth within limits
of good transient response and linear gray scale.
3. Control Track: Recorded 240 -Hz level, edit -pulse level and phasing,
and track width.
4. Audio Head: Track width, accuracy of placement, and gap- azimuth
alignment.
5. Audio Signal: Recorded signal level within limits of distortion com-
ponents, pre- emphasis, and audio bandwidth.
6. The final and most exacting step is interspliceability.
In addition to all of the preceding factors, the control -track signal phas-
ing relative to the reference pulse (hence recorded video tracks) must be
standardized. A spliced -in section which is different in control -track phas-
ing relative to the preceding and succeeding sections will need to be man-
ually retracked upon entering and leaving the spliced portion. Even under
the most fortunate operating conditions, a disturbing discontinuity will

occur.
In order to provide a broad background for the more detailed operat-
ing techniques, this section will consider the entire system concept of op-
erations. Following sections will detail actual operating procedures. It is
advisable for the reader to review this section after studying the remain-
der of Chapter 11. Studying in this way will make the chapter contents
most effective.
The built -in tape- monitoring oscilloscope will be used as a standard for
judging system setup and performance characteristics. Regardless of the
type of cro in your particular machine, the Tektronix Type 529, which is
popularly used in both the RCA TR -70 and Ampex VR -2000, is a good
place to start our discussion.
First, get a clear concept of what you are going to observe. The machine
has an "E -E" mode (review Fig. 5 -1, Chapter 5) which is quite important
in checks and setups, even though this is strictly "back -to -back electronics"
with the machine in the stop mode. Here is the basic idea to retain:_ The
E -E circuitry will show signal continuity, levels, and frequency and tran-
sient es-ponse through the video input amplifier, modulator, demodulator,
any time -base-error video amplifiers used (minus the correcting error sig-
nals, which must act while the system is in _playback), and the output
processing amplifiers. System sections and signals you don't see with E -E
circuitry are the record amplifiers, the fm signal being laid down by the
video head in the record mode, the playback preamplifiers, the channel
amplifiers (including equalization circuitry) , and the switcher output.
You should have the E -E circuitry in the best possible operating condi-
tion and then make recording and playback adjustments to give a match
with the performance of the E-E circuit.
See Fig. 11 -1A. This is the multiburst signal as it should appear at the
system input. Note that the sine-wave bursts are no longer reduced to
70 percent of white -flag level as was shown by Fig. 5 -2. (We are speaking
now of the latest high- band /low -band systems, not the older systems. If
you are concerned with an older system, keep the proportion of Fig. 5 -2.)
When the selector switch is in the input position, the monitoring cro is
actually "looking at" a point after the input -level control in the video
input circuitry. This enables you to adjust the input level to a standard
1 -volt composite amplitude, regardless of the level being fed from the
source. Also in later systems, there may be a high- frequency compensation
control as well as an input-level control. This enables you to set the input
frequency response for a "flat" multiburst pattern.
At this point let us emphasize that you have not established a standard
by which to judge anything! How do you know what any test signal looks
like at the actual input to the tape system? You must occasionally check
this with an external standard, which, in this case, means an external cro.
And if you use a sine -squared pulse test, you cannot use the Tektronix
--100
1111111 11111110
1/111111
1111111111II
I
11
II I II
111111
.;
I
Ail 1:
111111111111"' 1__._.__"' °" i
11111111II1 II 11I . 111i11 I
111111Í 11111 I'IIIÍ 11111 WAIL,. 1 1
1111111111111110 11 I111111110,1,11
1' l
11111 II111II I I
1
II1111111179Ih Ill
!;0
II;Ij 1' Ili I Il 1 Ili Ili '!i, ,i
Limits - nner: K 2% Outer: K - 4%

I
r-p Window and 0.125 us T -Pulse Response
Courtesy Tektronix, Inc.

(A) Flat position.
100 A
I I
Omits- Inner: K 2% Outer: K 4%

Window and 0.125 ps T -Pulse Response

(B) IEEE position.
- 100
80
Limits-Inner: K- 2% Outer: K -4%
Window and 0.125 ps T-Pu Ise Response
(C) Low -pass position.

Fig. 11 -1. Multiburst
Type 524 scope. This should be made evident as the discussion continues.
The recommended Tektronix scopes for modern systems and test tech-
niques are Types 545B, 547, or 453 with Mod 127C.
Thus, the "external standard" will be a scope with a 35- to 60 -MHz
bandwidth. Use this at periodic intervals to observe the terminated input
signal to the monitoring cro. The two should show the same response.
Use the external cro periodically to check the test signal input (actual
system input) across the input termination. This should be done for all
signal paths feeding the tape system and is particularly important for
color tape systems.
Use the multiburst test signal for all these checks. Then switch the
monitoring cro to the output and check the response of the E -E cir-
cuitry. On modern high -band recorders, you should see exactly what you
see at the input. On modern recorders in the low -band E -E mode, you
will normally see the last burst (4.2 MHz) slightly rolled off, with a
predominant "beat" pattern (moire) .
20 4 1 1
Limits -Inner: K- 2% Outer,* 4% _

Window and 0.125 us T-Puise Response
40

(D) High -pass position.
White Burst Burst Burst Burst Burst Burst

1 2 3 4 5 6
100 IEEE
0.5 1.5 2.0 3.0 3.6 4.2

MHz MHz MHz MHz MHz MHz
7.5 IEEE
40 IEEE
(E) Applied signal.

responses of waveform monitor.
You should check the remaining response positions of the cro, as shown
by Figs. 11 -1B through 11 -1D. The low -pass position (Fig. 11 -IC) is
quite important in the following respect. Only the dc axis remains. If
there is any "selective clipping," the axis will shift at the burst frequency
where this occurs. "Selective frequency clipping" simply means that clip-
ping occurs at one or more particular frequencies. When one side of a
sine wave is clipped and then the waveform is restored to a sine wave
by filtering circuits, the dc axis is shifted. The pattern of Fig. 11 -1C, as
well as all other patterns (except Fig. 11 -1A in low band), should look
the same at the output as it does at the input. (This still concerns the
E-E mode.)
NOTE: This chapter is not intended to give solutions as to what to do in

case of troubles. It is meant only to provide the operator with the proper
background to recognize when maintenance is needed. Maintenance is
taken up in Chapter 12.
Fig. 11 -1E shows the proper proportion of the multiburst amplitude

levels. Remember that when you calibrate the scope for 140 IEEE units
equal to 1 volt, then:
Volts/Unit = 1.0/ 140 = 0.00714

0 to 100 IEEE Units = 0.714 Volt, Blanking to Peak White
0 to -40 Units = 0.286 Volt, Sync Region
Now we will consider the sine-window test signal.' It is used because

it provides a simultaneous measurement of three characteristics which
should be grouped together into one K- factor for evaluation of system
performance. Fig. 11 -2A shows a continuous sine wave (top waveform) at
a frequency of 4 MHz. One cycle of this wave has a duration of 0.250 µs.
If the sine wave is shifted 90 °, there is one complete cycle of a 4 -MHz
cosine wave, starting and finishing at its negative peaks. With an added dc
component (lower waveform) to raise the negative peaks to the zero axis,
the T -pulse for a 4 -MHz system is formed. The half -amplitude duration
(h.a.d.) is 0.125 µs, or 1/8 µs, or one picture element in a 4 -MHz sys-
tem. The test signal generator normally supplies three selectable pulse
durations, as shown in Fig. 11 -2B.
Fig. 11 -2C shows that the significant energy spectrum of the T -pulse
is 50- percent (6 dB) down at 4 MHz, and there is practically no energy
beyond 8 MHz. The 2T -pulse is 50- percent down at 2 MHz, with prac-
tically no energy beyond 4 MHz. The T/2 -pulse is not used for per-
formance evaluation in vtr systems.
'For a more complete development of the sine pulse, see Television Broad-
casting: Systems Maintenance by Harold E. Ennes (Indianapolis: Howard W.
Sams & Co., Inc., 1972), p. 77 ff.
-0.250 ps = 4 MHz -
(A) Derivation of pulse.
DC Axis
0.125 vs = 1 Picture
Element
a.
172 =0.063ís= 1116us

T= 0.125Ús= 118{ís
21 0.250 Ds1/4Ús Frequency (MHz)
(B) Half-amplitude duration. (C) Energy spectrum.

Fig. 11 -2. The sine -squared pulse.
This energy spectrum provides a standard for system evaluation, much

the same as the standard VU meter was provided for aural gain riding.
This is to say that a known frequency spectrum is used. (The square -wave
response, formerly used, depended entirely on the rise time of the wave-
form and was difficult to interpret for transient response.) The window
signal has exactly the same rise time as its associated sine -squared pulse.
It is necessary to be sure you know how to establish and use this stan-
dard. First, study the scope evaluation in Table 11 -1. Note in particular
that the Type 529 waveform monitor exceeds the specifications of the
Type 524 scope to 8 MHz. An external standard must be much better
than the monitoring scope to set up and maintain the vtr system properly.
Actually, the Types 453 and 547 are the preferred scopes for a studio
system. Now note that although the T/2 -pulse is not used for the vtr, it is
used for checking the specifications of the Type 529 waveform monitor.
The T/2 -pulse has 50- percent energy at 8 MHz. Theoretically, the scope
which measures the pulse should have its -3 -dB response at least five
Table 11 -1. Tektronix Scope Characteristics

Response (to -3 dB) Rise Time
Type 453 -TV Scope

20 mV to 10 V/Div Dc to 50 MHz 7 ns
10 mV /Div Dc to 45 MHz 7.8 ns
5 mV /Div Dc to 40 MHz 8.75 ns
Type 545B
With lAl Preamp
50 mV /cm Dc to 33 MHz 11 ns
Type 547
With lAl Preamp
Type 529 Waveform Mon
50 Hz to 6 MHz Within 0.1 dB
50 Hz to 8 MHz Within 0.3 dB
Transient Response: 1/2 Pulse at Least 94 IEEE Units to Window
Midfreq Response: Within 1/2 IEEE Unit on Window
times better; this is 5 x 8, or 40 MHz. If the external scope shows the

T/2 -pulse at 100 percent of window height, the Type 529 monitor should
show at least 94 IEEE units with 100 units of window signal.
The sine-pulse measurement through the system under test is made in
terms of the first and second lobes (Fig. 11-3A), the amplitude ratios of
-h.a.d
(A) Lobe designations.
2nd Lobe
(B) Increasing delay with (C) Decreasing delay with

increasing frequency. increasing frequency.
Fig. 11 -3. Sin' pulse with ringing.
the leading -edge and trailing -edge lobes, the h.a.d., and (with the com-
bination window and pulse) the relative heights of the pulse and window.
A rapid rolloff (almost a cutoff) just above the video bandwidth con-
cerned results in practically no effect on the amplitude, but it does pro-
duce ringing. The shape of the rolloff and whether the resulting phase
shift is leading or lagging is revealed by the distribution of ringing before
and after the pulse. See Fig. 11 -3A. A fairly rapid rolloff with phase equal-
izers incorporated to correct the resulting phase distortion will cause the
ringing amplitudes to be equally distributed before and after the pulse.
The degree and nature of the high- frequency phase distortion is measured
by the departure from this optimum response characteristic. In Fig. 11 -3B,
we have an indication of increasing delay with increasing frequency. In
Fig. 11 -3C, we have an indication of decreasing delay with increasing
frequency.
The system performance K- factor is related to the following (simul-
taneous) characteristics:
High Frequency: K- factor of pulse

Low Frequency: K- factor of bar (window)
Midfrequency: K- factor of pulse -to -bar amplitude ratio
Fig. 11 -4A represents the graticule of the Tektronix Type 529 scope
with the sine- window signal displayed with normal horizontal sweep. The
limits for 2- percent K- factor are shown by dash lines, and solid lines indi-
cate the limits for 4- percent K- factor. The limits for the pulse-to -bar am-
plitude ratio are indicated at point 1. The limits on "tilt" of the bar (win-
dow top) are indicated at point 2.
Fig. 11 -4B represents the same signal displayed at 0.25H /cm with the
x5 multiplier. Note that the base line is at the 7.5- percent line, indicating
standard 7.5- percent setup.
Fig. 11 -4C represents the proper display of the pulse (only) to mea-
sure h.a.d. and any ringing amplitudes. The h.a.d. point (1) is indicated
by the termination of the K- factor lines, accounting for the 7.5- percent
pedestal used. When looking at the system output on the cro, remember
to adjust the pedestal output so that the base of the pulse is at 7.5 IEEE
units. Although the graticule is marked for a T- pulse, the 2T -pulse is
measured on the same graticule by proper adjustment of the time base.
With modern high -band systems, you should see no difference on the
Type 529 monitor at the input and output for a 2T -pulse in the E -E cir-
cuitry. The T -pulse will show a difference between the output and the
input. If proper phase correction is employed, an indication something
like that of Fig. 11 -3A will be obtained. The T-pulse will be somewhat
lower in amplitude than the window.
High- frequency rolloff results in loss of amplitude of the pulse. This
results in a widening of the h.a.d., since the area of the pulse represents
-40
2% and 4% K Factor 0.125 µsec T Pulse
(A) Display with normal sweep.
100
80
60
40
2% and 4% K Factor 0.125psec T Pulse
(B) Sweep set for 0.25H/cm X 5.
Fig. 11 -4.
a constant dc component. The energy is simply redistributed. A slow roll-

off within the energy spectrum produces a large reduction in amplitude
(and increase of pulse h.a.d.) with little or no ringing. A rapid rolloff
close to the top of the energy spectrum produces both a reduction ( perhaps
slight) in amplitude, and ringing. Ringing occurs at the frequency at
which the gain dip occurs in the system being measured. The amplitude
of ring depends on the sharpness of this gain -dip characteristic.
100
ao
60
40
20
1 1 1
20 I I + I
{ 1
}
f I
I
-40
2% and 4% K Factor 0.125 µsec T Pulse
For 2T Pulse: 0.250H/ cm x 25

For T Pulse: 0.125H /cm x25
(C) Display of pulse only.
Ringing_
Period
(D) Ringing period.
Waveform displays.
The ringing period (Fig. 11 -4D) has the following relationship to the
cutoff frequency:
R= f 1
where,
RP is the ringing period in microseconds,
fc is the cutoff frequency in megahertz.
For example, if the cutoff frequency is 4 MHz, the ringing period is:
RP = = 0.250 to
4
Then to find the cutoff frequency (in megahertz) for a given measured
ringing period (in microseconds):
f= R 1
Test Tapes (Low Band)

One of the most convenient methods of overall playback evaluation on
an operational basis is provided by the use of the SMPTE alignment tape,
based on Recommended Practice RP10. It should be noted that the manu-
facturer records the standard alignment tape according to the SMPTE
specifications; the SMPTE does not manufacture the tape. The customer
obtains his alignment tape from the manufacturer of his particular system.
In Signal Specifications for a Monochrome Video Alignment Tape for
2 -Inch Video Magnetic Tape Recording (SMPTE Recommended Practice
RP10), which follows, sequential bands from top to bottom of the raster
(each band of 16 to 17 lines representing a single pole -tip sweep across
the tape) are identified by numbers 1 through 16. The first band of lines
after that which contains vertical sync is identified as band 1; therefore,
band 1 is at the top of the raster. Fig. 11 -5 correlates the video -head num-
ber with the SMPTE band number. This correlation assumes that the cap-
stan tracking is adjusted so that the same head that plays vertical sync
also records vertical sync (head 1 for RCA and head 4 for Ampex) .
NOTE: Although this test tape is specified as being for a monochrome
recording, useful color evaluation is also possible. A high -band color test
tape is described later.
SMPTE Recommended Practice RP10

The material in this subsection has been taken from SMPTE Recom-
mended Practice RP10. For convenience, the paragraph numbering has
been retained.
1. Scope
1.1. This recommended practice specifies the signals to be recorded on
a magnetic video tape for use in evaluating and adjusting the perform-
ance of monochrome video-tape recording and playback equipment on a
routine operational basis. The characteristics which can be checked pri-
marily are related to the video performance, although a cursory check of
the audio channel is included for operating convenience.
Approx 8 Active Lines With Tracking Control Adjusted to Play Head Number SMPTE
Fallowing Vertical Blanking Vert Sync on Same Head as Recorded
/////////////////////////////
Band
RCA Ampex Number
///////////////////////////////////////// 2 2 1
3 3 2
4 1 3
1 4 4
2 2 5
3 3 6
4 1 7
Raster 1 4 8
2 2 9
3 3 10
4 1 11
1 4 12
2 2 13
3 3 14
4 1 15
' %/% ßíi .,, . .- '.:-;-",-.<7.7.,-.47,-......,

i 1 4 16
J
Start of Vert Band 16 Not Considered as Contributing Vert

Blanking to Active Picture Information Sync
Fig. 11 -5. Raster makeup related to heads and SMPTE band number.
1.2. Specifically, the recorded signals on the tape provide a means for
checking the following characteristics or adjustments:
(A) Video -head quadrature

(B) Tape vacuum-guide position
(C) Video levels
(D) Video amplitude- frequency response
(E) Video transient response
(F) Video low- frequency tilt
(G) Video amplitude linearity
(H) Video -head playback sensitivity
(I) Relative noise banding
(J) Rf- carrier deviation frequencies
(K) Program and cue-track audio levels
(L) Control -track levels and phase.
2. Recorded Signal Characteristics

2.1. The video signals recorded by the video heads shall occupy sequen-
tial bands from top to bottom in the reproduced picture, each band cor-
responding to a single traverse of a video head across the tape. For the
purpose of identification, these bands are designated as 1 through 16. The
first band after that containing the vertical synchronizing -pulse interval
shall be designated as band 1. (Band 1 will contain fewer active lines
than the other bands, because it contains a portion of vertical blanking.)
The active picture portion of the horizontal scan shall be divided into
11 equal sections. For the purpose of identification, these sections are
designated as 0 through 10. Information shall be recorded as follows:
2.1.1. (Bands 1 through 4) A stairstep signal consisting of a 10 -step
linear gray scale extending from blanking level to 100 IEEE units, re-
spectively (Fig. 11 -6A) .
2.1.2. (Bands 5 through 8) A stairstep signal consisting of a 5 -step

linear gray scale extending from black level to 50 IEEE units, respectively
(Fig. 11 -6B) .
2.1.3. (Bands 5 through 8) A series of 5 sine -wave bursts (Fig. 11-6B)
described as follows: The time sequence of the burst frequencies shall be
4.2, 3.6, 3.0, 2.0, and 1.5 MHz. The axis of the multiburst shall be at 30
IEEE units, and the peak -to -peak amplitude shall be 40 IEEE units. Each
burst duration will be at least 75 percent of the section width.
2.1.4. (Bands 9 through 15) A window signal at reference white level
(100 IEEE units) 3 sections wide and 6 bands high to be positioned hori-
zontally in sections 6, 7, and 8 (as shown in Fig. 11 -6C) and vertically
between the centers of the ninth and fifteenth bands. The remaining sec-
tion shall be at blanking level (0 IEEE units) .
2.1.5. (Band 16 only) Vertical synchronizing -pulse interval and a por-
tion of vertical blanking.
2.1.6. (Bands 1 through 15) Sine -squared pulses %s microsecond wide
(measured at half level) and 50 IEEE units high at horizontal positions
corresponding to the center of each of the first six sections. The base level
of each sine -squared pulse shall be as follows:
(A) Bands 1 through 8, the same as the accompanying stairstep
section level, as shown in Figs. 11 -6A and 11 -6B.
(B) Bands 9 through 15, at blanking level, as shown in Fig. 11 -6C.
2.2. The waveform of the composite signal shall appear as shown in
Fig. 11 -6D.
23. All synchronizing waveforms and signal amplitudes shall conform

with EIA Standard RS-170 or the latest revision thereof.
2.4. All video signals shall be within ±1 IEEE unit of specified am-
plitudes.
2.5. The leading and trailing edges of the window signal shall corre-
spond in shape and rise time to the sine -squared pulse specified in para-
graph 2.1.6.
2.6. Overshoot of the stairstep signal shall not exceed 5 percent of the
amplitude of transition. An exception is the trailing edge of the stair -
step (leading edge of horizontal blanking), which is limited to 2 percent
in accordance with EIA Standard RS -170 or the latest revision thereof.
2.7. Multiburst frequencies shall conform with specified values within
1 percent. Total harmonic distortion content of the multiburst frequencies
shall not exceed 2 percent.

2.8. The audio tone and cue records shall consist of an audio tone inter-
rupted periodically with voice announcements.
2.9. (A) The audio tone shall be 400 Hz ± 2 percent recorded at a
level 10 dB below that corresponding to a 3- percent total harmonic dis-
tortion at 400 Hz. (B) The audio response- frequency characteristics shall
be as specified in "Proposed American Standard Characteristics of the
Audio Records for 2 -In. Video Magnetic Tape Recordings" (VTR 16.5)
or the latest revision thereof.
2.10. The voice announcements shall be made at 1- minute intervals and
shall not exceed 20 seconds in duration. The announcement shall provide
identification of the tape as regards the applicable SMPTE recommended
practice, the tape issue number, and the manufacturer of the standard
tape. Additional identification (such as serial number) may be included
at the discretion of the manufacturer.
3. Recording Conditions
3.1. The video- alignment tape shall conform with applicable American
standards and SMPTE recommended practices.
Use of the Test Tape
Special care and precautions should be taken in the handling of an
alignment tape. The following quotes are from the RCA instructions for
the MI -40793 alignment tape (SMPTE) :
"Since the tape can be affected by changes in temperature and humidity,

extreme changes are to be avoided. The following atmospheric conditions
are recommended for the area in which the tape is stored:
Relative Humidity: 40 -60 Percent
Temperature: 60-80°F
"Do not attempt to use the tape until it has been stabilized for at least
16 hours at room temperature.
"Take precautions to avoid erasing the tape accidentally. Do not place
or store the tape in the region of magnetic fields.
"Make certain that the heads and guides on the recorder are clean be-
fore using the tape.
"Check the setting of the reel brakes. Rapid stops with improperly set
brakes can damage or distort the tape.
"Be careful to wind and rewind the tape properly; the edge of the tape
must not climb up the side of the reel.
"After long usage, short-duration, small pressure changes on the order
of 0.0001 inch (0.1 mil) may begin to appear in the tape during play-
back. However, the tape may still be used by choosing the average vacuum -
guide position."
A photograph of the monitor display of the input signal for the SMPTE
tape at the time a recording is being made is shown in Fig. 11 -7. The
bands and sections established by Recommended Practice RP10 are identi-
fied on the photograph for convenience. Fig. 11 -8 is the oscilloscope dis-
to
100- Stairsteps
9
8
80- (
7 I
60- Sine- Squared Pulses
40-
20-
-Section 0 Section 10-- -i

-20-
40 Active Picture Scan
(A) Bands 1 through 4.

100 -
80 -
Sine -Squared Pulses Stairsteps
60 - 4.2 MHz 3 MHz 5 MHz
40
20
3.6 MHz 2 MHz

o-
-20 - - Section 0 Section 10
40 - Active Picture Scan
(B) Bands 5 through 8.

Fig. 11 -6.
play (horizontal -rate sweep) of the monitor signal in Fig. 11 -7. It is

obvious from Fig. 11 -7 that adjustments for quadrature and vacuum -guide
position may be made to eliminate geometric distortion of the vertical
lines, as previously described. Most of the remaining system evaluation is
obtained by analysis of the oscilloscope waveform display. Photographs of
the output signals taken from an RCA TRT -1B recorder while playing the
alignment tape are shown in Fig. 11 -9. Since the signals are keyed only
onto definite portions of the composite signal, the line -selector trigger
must be used on the oscilloscope. By selecting the appropriate sweep
speed, any particular line or lines can be displayed. Normal horizontal -
rate sweep is used for the composite signal display shown in Fig. 11 -9A.
When you have checked the E -E circuitry for proper characteristics, play
the SMPTE alignment tape, and observe as closely as possible the T -pulse
100-
WO -
^
Window
80-
Sine- Squared Pulses
60-
40-
20-
0-
-Sections 6, 7, 8--
-20-
-40- Active Picture Scan
(C) Bands 9 through 15.

100-
Window
80-
Stairsteps
Sine-Squared Pulses I
60-
40-
20-
O-
1-Section 0 Section 10
-20-
-40 - Active Picture Scan
(D) Composite waveform.

SMPTE waveforms.
response in the playback mode. (This tape at present uses the 0.125 -µs
T-pulse.)
Then it is a simple matter to compare the record mode with the play-
back mode. Make a recording of the T -pulse and window signal from a
properly set up test signal generator. On the latest high -band color sys-
0 I 1 2 3 4 5 6 I 7 8 9 I 10
Sections
Courtesy RCA
Fig. 11 -7. Monitor display of SMPTE input signal at time of recording.
tems, the overall performance should be better than the present SMPTE
RP 10 alignment tape, which is a low -band monochrome test.
The playback waveform should be within the 4- percent K factor for
the T -pulse (on modern systems) . The actual specifications are normally
given in terms of the 2T- pulse, where the playback should be well within
Fig. 1 -8. Oscilloscope display of

1
SMPTE input signal at time

of recording.
Courtesy RCA
(A) Composite waveform. (B) Bands 1 through 4.
(C) Bands 5 through 8. (D) Bands 9 through 15.

Fig. 11 -9. Playback of SMPTE waveforms.
the 2- percent K rating. This becomes particularly important when dubs

are to be made from a master.
Test Tapes (High Band)

At the time of this writing, only one manufacturer (RCA) is recording
a high -band test tape which essentially follows SMPTE Recommended
Practice RP -43 -1971. Ampex is working on such a test tape, which prob-
ably will be available by the time this book is in print. Following is a
description of the RCA high -band test tape.
The RCA high -band alignment tape, MI- 41699, is a reel of 2 -inch video
tape on which has been magnetically recorded a series of signals to facili-
tate the adjustment and performance verification of quadruplex video tape
recorders. The tape has a minimum playing time of 10 minutes and is in-
tended for use with television systems operating on 525 -line, 60 -Hz NTSC
standards. The signals on the tape are arranged to facilitate the use of
the tape either for routine operational checks or for more detailed analysis
of recorder performance. A 1000 -Hz tone has been recorded at 8 dB below
the 3- percent distortion level and may be used to establish the correct
record level.
The signals recorded on the tape provide the means for checking the
following characteristics and adjustments:
(A) Video -head quadrature

(B) Vacuum -guide position
(C) Video levels

(D) Comparison of carrier frequencies of the video recording system
(E) Verification of level and phase of the control -track recording
system
(F) Indication of video frequency -response characteristics of the re-
producing system
(G) Video amplitude linearity
(H) Video transient response
(I) Chrominance-luminance delay
(J) Program audio level
(K) Indication of tape speed
The use of the alignment tape will allow the adjustment of the vacuum
guide to the position which will assure interchangeability between tapes
made on correctly adjusted equipment.
The contents of this tape, and hence its proper usage, may be best under-
stood by referring to Fig. i1 -10. The contents of the tape are shown sche-
matically, and brief notes regarding the intended use of the tape are given.
This drawing should be referred to from time to time by the reader as he
proceeds through the following paragraphs. (Certain of the waveform
photos are included in Chapter 12.)
The tape is divided into two major sections. In the first section, the
various alignment signals are displayed in fairly short -duration groups.
This section of the tape is intended for quick cursory examination of
recorder performance, and can be used by an operator in performing a
routine check at the beginning of the day's operation or just prior to
making a recording. The second part of the tape presents the same signals,
but in segments of much greater duration. This section of the tape is in-
tended for meticulous adjustment or correction of faults in the recorder
(Chapter 12) .
Since the tape can be affected by changes in temperature and humidity,

extreme changes are to be avoided. In the area in which the tape is stored,
the relative humidity should be in the range of 40 to 60 percent, and the
temperature (Fahrenheit) should be between 60 and 80 °. In addition,
the following precautions should be observed:
1. Do not store or place the tape in the region of magnetic fields.

2. Do not attempt to use the tape until it has stabilized for at least
16 hours at room temperature.
3. Be sure that the heads and guides on the recorder are clean before
using the tape.
4. Check the setting of the reel brakes, capstan pinch roller, and reel
tensions.
5. Use a take -up reel having a diameter of 8 inches (same size as test -
tape reel) to maintain a proper inertia balance between reels.
Start Black
0:00
0:20
Multiburst
0.50
Ramp + Subcarrier This sequence of signals is to permit video head
1:15 alignment and guide alignment, and to provide a
Sint Window 2T, 201 qualitative check of the playback system.
1:40
Color Bars
2:10 Tape to be stopped or rewound in this black segment
2:20 only, if the complete recording is not to be played back.
Multiburst To be used to indicate video level and frequency -

response settings.
3.50
To be used to measure differential gain and

Ramp + Subcarrier
differential phase.
6.20
To be used tocheck low- frequency response, transient

Sin2 Window 2T, 201 response, chroma amplitude, and envelope delay near
the subcarrier frequency.
7.50
To be used to check the response of the system to

Color Bars -75% color bars, and to provide a final subjective
test of playback performance.
10:20 Tape to be stopped, or rewound from this segment.

10:30
Black
End
Video signals consist of composite sync with indicated test signal.
The tone recorded on the audio track is 1000Hz at 0dB.
Fig. 11 -10. Diagram of test tape with associated aural identification.
6. Stop or rewind the tape only in the black -signal portions between
recordings. Rapid stops, with improperly set brakes, can damage or
distort the tape. Observe the tape as it winds on the reel. The edge
of the tape must not climb up the side of the reel, or permanent
damage may result.
7. Take precautions to avoid erasing the tape accidentally, i.e., by
pressing a setup or record button.
NOTE: After long usage, small apparent guide- pressure changes on the
order of 0.0001 inch may begin to appear in the tape during playback.
However, the tape may still be used by choosing the average guide posi-
tion.
Errors in quadrature of the video headwheel can be measured with the

alignment tape. This tape was made with precision quadrature (30 nano-
seconds peak to peak maximum) When a headwheel with errors in
.
quadrature is used to play the precise tape, horizontal displacements di-

rectly proportional to the errors in quadrature will be visible in the repro-
duced picture at the output of the demodulator. These errors will be
corrected by the automatic timing corrector, and their magnitude may be
measured by observing the error -signal display if the sensitivity is known.
In RCA machines, the sensitivity of the display has been set so that 100
IEEE units represent 1 microsecond of error. On RCA high -band head-
wheel panels, the quadrature has been set very precisely by the manu-
facturer. It is not recommended that the adjustment of quadrature be
attempted in the field.
The correct reproduction of the video signal indicates correct position-
ing of the vacuum guide. The effects of playback with an incorrectly posi-
tioned vacuum guide can be separated into two basic types of errors. Com-
binations of these errors usually occur. An error in vacuum -guide position
parallel to the plane of the tape is called guide height error and results in
the reproduction of vertical bars as a series of scallops or bows. An error
in vacuum -guide position perpendicular to the plane of the tape is called
penetration error and results in the reproduction of vertical bars as a
series of zig-zag lines called skewing or jogs (Chapter 2 ) .
The position of the vacuum guide of the television recorder must be

adjusted until the vertical bars of the alignment tape are reproduced as
straight vertical bars with minimum horizontal displacements. Adjust the
guide height to remove scallops, and adjust the guide penetration to elimi-
nate skewing. Here again, the ATC error signal may be used as a sensi-
tive indicator of the magnitude of the errors, and the guide should be
adjusted to minimize these errors (described later) The position of the
.
vacuum guide should be rechecked at 20 -hour intervals throughout the

life of the headwheel.
Before proceeding with the alignment procedure:
1. Rotate the guide -position control fully counterclockwise for mini-

mum penetration.
2. Observe the output of the demodulator on the picture monitor.
3. Place the recorder in the play mode of operation.
4. Slowly adjust the manual vacuum -guide control knob to minimize
vertical misalignment of the picture elements and to avoid over -
penetration.
5. Do not stop the tape or rewind except at the point 2 minutes and
10 seconds from the beginning or at the end during the black sig-
nal (see Fig. 11 -10).
6. Proceed with final guide alignment according to the procedure out-
lined later, and remember to avoid overpenetration.
QUADRUPLEX TAPE SYSTEM OPERATIONS D,35.
WARNING: Do not stop, initiate forward wind, or rewind in any por-
tions of the tape other than black. Otherwise the active signal portions of
the tape may be stretched. Stretching of the tape will produce distortion of
the tape and effectively result in guide- position errors on successive replays.
The o rational use of this test tape will become evident as this chap-
ter progresse se of the tape in maintenance is covered in Chapter 12.
.
Use of Color rs
The pro er roportionment for a 75-percent color -bar signal is shown
in Ig. -11A. Note that this incorporates a 7-percent setup. The gray
100 - 100 100
89
80 - 77 77
72
69
.15
60 - 56
48
Tolerance . 2.5 IEEE Units 46
en Dk -pk Chrominance Levels
d 40 - 38
36
28
20 - 15
12
7
0- 1
-5
-20 - The duration of each of the bars is

117 of the scan between blankings.
-16 -16
Tolerance « 10%
40
l
(A) 75% bars, not split- field.
s 53'
120
Ref White.
80
60
40
22
20
Rel Black
0
(B) 7 5°70 bars, split-field.

Fig. 1 1 -1 1 . Color -bar signals.
bar then, instead of being 70 percent, is properly peaked at 77 percent

because of the 7- percent setup. You should adjust the input level so that
(providing you have properly adjusted the color -bar generator) this level
occurs on the first (gray) bar. Then adjust the input high- frequency com-
pensation (when used) to bring the tops of the yellow and cyan bars to
the 100 -percent level. This drawing is in detail for the setup discussion
later.
The split -field color -bar pattern is shown in Fig. 11 -11B. In this case,
the white reference is at 100 percent, and the gray is still about 77 percent
of peak white (with 7-percent pedestal). Again, adjust the input high -
frequency compensation to just fill in the white reference with the chroma
of yellow and cyan. You can place the scope on IRE response (Fig. 11-
12A) to adjust the white reference to 100 percent. Then with the scope
on normal response (Fig. 11 -12B), adjust the chroma level to fill in the
white reference pulse.
White Reference Pulse
(A) Scope on IRE response. (B) Scope on wideband response.

Fig. 11 -12. Split -field color -bar patterns at horizontal rate.
Sometimes, you may encounter a tape with a leader that has color bars
at 100 -percent amplitude. Thus the yellow and cyan peaks are at a level
33 percent over the white reference. This can be confusing on the straight
color -bar pattern of Fig. 11 -11A and cause you to set the wrong playback
level unless you can recognize proper setup. You should always use 75 -per-
cent color bars, not 100 -percent bars. The advantage of the split -field color -
bar pattern (Figs. 11 -11B and 11 -12) is that a white "flag" always exists as
a 100 -percent reference level, to indicate clearly any difference in lumi-
nance and chrominance gain.
First, become familiar with the cro display of your own station color -
bar signal. Looking at the input with horizontal-rate sweep, you should
see the waveform of either Fig. 11 -11A or 11 -11B for a properly equalized
input.
Now observe the same signal with vertical -rate cro sweep. The top tips
of the chroma (yellow and cyan) will occur at 100 IEEE units (Fig. 11-
13). Note from the drawings of Fig. 11 -11 that the bottom peaks of
chroma (red and blue) should occur at -16 IEEE units and the burst
bottoms should be at -20 IEEE units.
Since the sweep is at the vertical rate, the chroma signal from -16 to
+100 IEEE units results in a rather heavy trace. The space between -16
and -20 IEEE units, since the burst duration is shorter than the other
signal intervals, results in a fainter trace. (The graticule illumination was
turned off for the photograph of Fig. 11 -13 so that the faint part of the
trace would be visible.)
A color tape should contain at least two minutes of color -bar recording
on the leader. This enables the operator to set the individual channel
equalizers properly for playback of the program content.
White Reference Pulse
-16 Units
I- -20 Units
Fig. 11 -13. Split -field color -bar pattern at vertical rate and
with wideband response.
Setting Playback Equalization

Review Fig. 11 -5 carefully. Output from the head which plays back
vertical sync (head 1 for RCA and head 4 for Ampex) will appear at
the extreme left of the vertical -rate cro pattern. Once each field, this head
delivers only a few active lines of picture information, so it provides only
about 3/4 of the information delivered by the other heads.
The second head (SMPTE band 1) also contains only about half of the
normal active picture lines during one pass in each field. So here is a good
procedure in setting up the individual channel equalizers on playback:
1. Look at the demodulator output with the scope at the horizontal rate.
Adjust the equalizer for the head playing back vertical sync (head 1
for RCA and head 4 for Ampex) so that the green bar (Fig. 11 -11)
is at the chroma level shown. Of course all of the individual head
contributions are being indicated, but you can recognize the indi-
vidual channel you are adjusting by watching the resulting (faint)
trace, within the green bar area, move on the scope pattern.
2. Then change to the vertical sweep rate (Fig. 11 -14) . Adjust the
remaining channel equalizers so that the chroma peaks are at -16

and +100 IEEE units. Do not fill in the space between -16 and
-20 units; this should remain a faint trace indicating individual
burst levels.
3. After all playback channels have been matched as closely as possible
on the cro, observe the output of the system on a good color monitor.
Trim the individual equalizers slightly for minimum color banding.
4. On the cro display of the demodulator output at the horizontal
rate, the burst amplitude should be correct (40 IEEE units). This is
the proper input to the signal -processing amplifiers. An attempt at
"overequalization" at this point can result in servo instability.
(A) Channel equalizers not (B) Equalizers adjusted for same

properly adjusted. chroma level.
Fig. 11 -14. Demodulator- output displays (wideband scope response).
5. Now change the cro to the system output and again check the color -
bar pattern at the vertical rate. In a tape system, the luminance -to-
chrominance ratio will be correct with proper equalization, but the
burst may not be. This involves a separate control which must be
set for proper output amplitude.
NOTE: You can also match all four head -equalizer adjustments by observ-
ing the output of the system on a vectorscope (where automatic chroma
equalization is not used or is removed for this adjustment). Fig. 11 -15
shows the appearance of multiple dots along each vector, showing un-
equal chroma amplitudes. Fig. 11 -17 shows the proper indication for
equal chorma amplitudes from all four heads (dots superimposed so that
a single dot appears for each vector.) Since each "dot" (vector) in Fig.
11 -15 is only 1/4 normal brightness, the trace is very dim. For this photo-
graph, the graticule line illumination was turned off, making the IEEE
scale more visible than the vector scale. When all dots are superimposed,
the trace brightens considerably, as in Fig. 11 -17.
Fig. 11 -16 is an example of how much improvement over the earlier

low -band, monochrome systems has occurred in modern tape recording
systems. Fig. 11 -16A is the appearance of the multiburst pattern on play-
Multiple Dots burst

Yellow Vector
Fig. 11 -15. Vectorscope display of color -bar playback with channel equalizers
not adjusted for proper match.
back with an early machine (RCA TRT -1A) when the playback head
equalizers were adjusted for an optimum picture. In these early systems,
it was seldom possible to match response at all discrete sine -wave fre-
quencies, especially on a tape -interchange basis. The best technique at
(A) Best reproduction on (B) Best reproduction, modern

older system. high -band system.
Fig. 11 -16. Monitor presentations of multiburst playback.
that time was to adjust all channel compensation controls for the best
match (no banding) on the first two burst frequencies ( normally 0.5
MHz and 1.5 MHz). Most of the average picture content of normal pro-
grams (in monochrome) is in the spectrum up to 1.5 to 2.0 MHz maxi-
mum. Thus, although a test signal such as the higher frequencies in a
multiburst pattern would show some banding, the usual monochrome pic-
ture content was satisfactory.
For a color signal, there is a very different situation. The chroma con-
tent lies in sidebands of the 3.58 -MHz subcarrier. Therefore, optimum
response up through the highest frequency of the multiburst pattern (nor-
mally 4.2 MHz) is extremely important. Fig. 11 -16B is a photograph of
the playback of the same multiburst pattern as that of Fig. 11 -16A, except
that it was recorded and played back on a modern high -band color record-
ing system. For a properly operating high -band system, it is necessary to
get extremely close to the monitor presentation to observe any "noise" at
all on the multiburst signal. On modern low -band systems, some slight
moire and noise will be visible on the last burst (4.2 MHz) of the
pattern.
Adjusting Monochrome and Color ATC Error Gains
Never play the monochrome and Color ATC error gains against each
other while watching a color monitor for transients and best "effect." If
you do, you may have trouble with a spliced -in portion or the next color
tape. There is really only one proper method for each:
(A) For monochrome time -error gain: Put the guide servo in the
manual position, and while observing the demodulator output on
the picture monitor, insert about a microsecond of error. Then
switch the picture monitor to the output, and adjust the mono-
chrome time -base -error gain to correct all geometric error ex-
actly. Return the guide servo to minimum error as observed at
the demodulator output.
(B) The color error gain should be adjusted while observing the
vectorscope on color -bar playback. Adjust this gain for minimum
jitter of the yellow vector. Yellow immediately follows the time
of correction, which occurs during horizontal blanking, and is
most influenced by the correction.
Observe the yellow- vector dots and the I and Q dots in Fig. 11 -17. Note
that these dots have wider excursions in Fig. 11 -17A than is true for
Fig. 11 -17B. You will note that the proper error -gain control setting is
most evident by observing only the yellow, I, and Q vector dots. If the
color -bar signal does not contain I and Q signals, observe only the dot for
the yellow vector.
Always recheck the burst -phase setting with the vectorscope after ad-
justment of the Color ATC error -gain control.
Yellow Burst -1 rQ Yellow Burst -I tQ

(A) Dot displacement (jitter) . (B) Adjusted for minimum jitter.
Fig. 11 -17. Optimization of Color ATC error gain control.
System Phase and Burst Phase

Be sure you understand the difference between "system phase" and
"burst phase" in color operation. See 11 -18A. The color error detector re-
ceives a reference 3.58 -MHz signal from the local generator and a burst
from the tape playback signal gated on by a horizontal pulse. If the phase
of the reference signal is changed in this path, the reference that is com-
pared with the tape- signal burst to obtain an error signal is changed. Hence
all the vectors are rotated, as in Fig. 11 -18B.
The color processor receives the same reference subcarrier, and also
signal chroma (sidebands) from separated high- frequency video informa-
tion. (The luminance information and chrominance information are sepa-
rated to provide individual clamping and processing.) If the phase of
the reference signal in this path is changed, the burst vector is rotated
relative to the signal chroma information, as indicated by Fig. 11 -18C.
It makes no difference to the color monitor or receiver what system
phase it "sees." It synchronously demodulates the color information in
accordance with the relationship between the burst and the picture side -
band information. So if you change the burst phase, you change flesh tones,
hues of color bars, etc.
You can vary system phase from one limit to the other, and the color
monitor will not change hue if it is connected to the output of the tape
system. This control must be used in matching the phase of the color tape
signal with other local color sources, as when special -effects equipment
must be used. The special- effects amplifier will pass only one burst source
at a time. Therefore, if you mix color tape with a local color signal, any
difference of system phase will result in mismatched colors.
To adjust system phase properly, feed the output of the special- effects
amplifier to a color monitor. Feed color bars (reference) to one input,
"wipe in" the color tape signal which also contains color bars, and rotate
the system -phase control on the tape system to obtain a match.
When you do not use special effects or other mixing equipment, your
only concern in phasing is burst phase. The best way to adjust this (other
than using a vectorscope) is to have a color monitor which can be
switched between the regular program -line output and the tape- system
output. When the monitor is adjusted for proper reproduction on the
program -line output, switch it to the tape- system output and adjust the
tape burst -phase control for proper color.
Tape- Signal
Burst
System Subcarrier
Reference
Subcarrier
Amp
\ Color Error
Detector
Gated Horiz
From Sync Logic
System Phase
1Phase Error
Burst Phase
Error Signal
Color
to Colortec or
Processor
Color ATC
Burst Subcarrier
1
Signal Chroma
Information
1Sidebandsl
(A) Block diagram.
.r //
i
Burst Burst
(B) System phase. (C) Burst phase.

Fig. 11 -18. Phasing of color vectors.
Normally this is done on the color -bar leader of the tape to be played.
It is nearly always necessary to readjust slightly for proper flesh tones in
the program content. The burst-phase setting is rather broad on color
bars (unless a vectorscope is used) compared to the much more subtle
phase adjustment required to obtain good flesh -tone reproduction on a
color monitor.
11 -3. PRELIMINARY SETUP PROCEDURE
At the start of the operating day, before the tape is threaded on the
transport, ample time should be scheduled for the operator to carry out
the following procedures. (The cleaning procedure should be exercised
before every recording and before playback of programs more than 5 min-
utes long. Once a day is normally sufficient for the other procedures.)
Cleaning
Naphtha, Energine, and Freon TF are the most widely accepted clean-
ing agents for video and audio heads and all other metal contacts on the
tape transport. Chlorothene may also be used, but only with the utmost
caution. This agent (although cheaper than Freon TF in bulk quantities)
will soften the oxide binder of the tape if ample time is not allowed for
it to dry thoroughly. Denatured alcohol is also used for all items other
than audio and video heads.
Use lint -free tissue (such as Kim -Wipes) or a soft brush (camel's hair)
moistened with the solvent to clean the following items:
A. Supply tension arm
B. Supply idler
C. Video (master) erase head
D. Rotating -head panel (headwheel or drum, pole tips, vacuum guide,
nylon bearings under vacuum -guide block, vacuum -guide screw
contact with the movable servo arm, control -track head)
E. Audio and cue heads (and simulplay audio and control -track heads)
F. Take -up idler
G. Capstan and capstan pinch roller
H. Counting roller
I. Take -up tension arm
J. Take -up and supply reel hubs
The video heads should be cleaned by hand -rotating the shaft while
holding a moistened tissue against the heads and rotating with a slight
circular motion. Use the agent on the tape -supporting surface of the
vacuum guide while the vacuum pump is working, and inject a little of
the agent into the slots with an eye dropper or hypodermic injector. In-
ject a slight amount into the slip -ring assembly (on RCA heads) on the
head panel, and rotate the shaft by hand to clean the brush assembly. In-
spect each brush to be certain that the slim spring contact arm is prop-
erly seated in the brush groove. A cotton applicator or tissue dampened
with a solvent is handy for cleaning the control -track head.
Degaussing
As a precautionary measure, the tape transport should be demagnetized
at least once daily. Fig. 11 -19 shows the usual tell-tale indication in the
switched rf output at the switcher unit when magnetic paths exist. A tape
which was originally good can be harmed by this condition.
NOTE: The pattern in Fig. 11 -19 is observed on the Ampex A scope

when the selector switch is in the SWR RF Output position. This moni-
tors the off-tape rf signal from each of the four video heads. It is also
as observed on older RCA systems at the FM Level or SWR Out moni-
toring position. The RCA TR -70 fm level pattern is different, as illus-
trated in Section 11 -6.
The small hand degausser which is normally purchased with the equip-
ment should be energized at least 4 feet from the tape transport and
brought slowly up to the panel. Keep the unit slowly rotating in a small
Fig. 1 -1 9. Oscilloscope presentation

1
of switcher rf output, indicating

a lY1NA wM1VIIIIuiAw1 magnetic path.
circle while going over the entire tape path from supply reel to take -up
reel; be sure to include the master video erase head. Start the video heads
rotating (push the OVERRIDE switch on the Ampex machine or the
STANDBY or SETUP switch on the RCA machine) as the degausser is
passed over the rotating-head panel. Degauss the audio heads with the
conventional audio-head unit. Slowly pull the degausser away from the
panel and de- energize after the unit is at least 4 feet away.
The fact that no magnetic tools, such as screwdrivers, wrenches, etc.,
should be stored near the transport or tape is often forgotten. Always de-
gauss tools before using them on the equipment or storing them in
nearby shelves.
11 -4. LEVEL CHECKS
Always check to see that the monitoring oscilloscope is properly cali-

brated before setting levels. The usual sequence is as follows:
1. Input: 1 volt (peak-to-peak)

2. Demodulator output: 1 volt peak-to -peak (This is also the input to

the ATC system and processing amplifier.)
3. Processing amplifier: 1 volt peak -to -peak with 5- to 10- percent setup
(interval between maximum picture black and blanking level) , and
0.4 volt from blanking to sync tip. These controls are all included
in the processing amplifier.
When it is certain that a 1 -volt peak -to -peak level exists at the input,
the deviation control for fm modulation should be set by the means pro-
vided in the system. For example, Fig. 11 -20 shows how the deviation
control is adjusted on the RCA TR-70.
11 -5. HEAD -ALIGNMENT AND SELF -CHECK PROCEDURES

The normal sequence of checks (and adjustments where required) is:
vertical (height) alignment of vacuum guide for maximum concentricity,
horizontal alignment of the vacuum guide for proper tip penetration of
the particular rotating head assembly, and electronic or mechanical align-
ment for quadrature (in older systems) These checks, and other self -
.
check procedures which are not a normal operational procedure unless

questionable conditions exist (such as a doubtful alignment tape or tem-
porary loss of such tape) , are described in the following paragraphs.
Vertical (Scallop) Adjustment
See Figs. 2 -17 and 2 -18 for monitor displays of an alignment tape when
the vacuum guide is too high or too low. When the guide is properly
set, the video head tips exert the same degree of tape stretching (result-
ing in the same relative velocity) throughout the entire arc of contact.
This indicates maximum concentricity of the vacuum guide with the arc
described by the pole -tip radius of rotation.
A. Thread the manufacturer's alignment tape, usually containing verti-

cal lines recorded under precise laboratory measurements, onto the
tape transport and place the system in the play mode.
B. Place the vacuum -guide servo in the automatic mode of operation
so that the tip penetration is somewhere near standard. NOTE: With
a new or rebuilt head assembly, a rough skew or jog adjustment
(next subsection) may need to be made before the height adjust-
ment. The need is apparent if the tips are obviously not contacting
the tape, or if the pole tips are obviously digging too deeply (over-
loading the servo system) In either case, the tape will not play.
.
C. While watching the monitor, turn the guide- height or scallop -

adjustment screw to eliminate any scalloping of the vertical lines.
The alignment points are shown in Figs. 2 -22 and 2 -24.
D. One method of rapid self -check without the alignment tape is to
use the principle that there is normally some slight eccentricity of

the vacuum-guide center of curvature with the axis of the rotating
heads. This means that as the guide is moved farther away from
the pole tips, contact is first lost at the center of the tape (Fig. 2-
13B) With the scope looking at an individual head playback -
.
channel output, back the vacuum guide away from the heads, by
means of the electrical tip-projection control with the guide servo
placed in the manual mode of operation, so that contact exists only
Reference Pulse
Video Same Height

as Reference Pulse
(A) Proper deviation.

Reference Pulse
_ Video Lower Than

Reference Pulse
(B) Underdeviation.
Reference Pulse
_ Video Higher Than
Reference Pulse
(C) Orerdeviation.
Fig. 11 -20. Demodulator output displayed at vertical rate.
at the edges of the tape. Fig. 11 -21 shows a typical pattern obtained
under these conditions. If the guide is perfectly concentric to the
pole -tip rotation (the same pressure at top and bottom of tape) , the
height of the pulses will be equal. Adjust the height- adjustment
screw until the pulses are as nearly equal as possible. Then return
the tip -projection (guide position) control to normal.
Fig. 11 -21. Single- channel playback

waveform with guide adjusted to
barely contact tape.
E. A more accurate method of self -check for height adjustment is to

make a recording consisting of vertical lines, such as a grating -
pattern generator signal, and simultaneously to feed the control -
track signal to the audio track. Let the tape run through to the take -
up reel and do not rewind the tape. Simply place the take -up reel
with the run -off tape onto the supply -reel hub and thread in the
normal manner. It is now obvious that the tape is reversed -the
audio track (which also contains the control track in this case) is at
the bottom instead of the top, and as the role tip sweeps across the
tape in playback the signal will be reversed from the recording
mode. Thus, playback of the tape will reveal scalloping if the
vacuum -guide height adjustment is improper. This self-check method
is accurate but time -consuming. Use the simplest nonsynchronous
playback mode for a more stable picture.
Horizontal (Skew or Jog) Adjustment

Horizontal adjustment is normally required only on the initial installa-
tion of a new or rebuilt head assembly. When the vacuum guide is once
set to the standard distance from the headwheel axis, the heads will record
and play back tape interchangeably over a wide range of head wear with-
out resetting.
A. Place the vacuum -guide servo in the manual mode of operation. With
the standard alignment tape threaded, place the system in the play
mode. Remove all high -frequency post- emphasis (where used) to
avoid possible geometric distortions of the vertical lines from trans-
ient effects on leading and trailing edges.
B. Set the guide -position (electrical) control on the control panel to
the position designated by the manufacturer for standard.
C. Adjust the penetration -range adjusting screw on the head panel for
straight vertical lines. (See Figs. 2 -14 and 2 -16 for time errors result-
ing from improper horizontal positioning of the guide.)
Keep the time -base concept of this velocity error in terms of the picture -
signal components. This will serve well in the future. We will discuss now
one practical application, determining pole -tip projection remaining on a
head.
Thread the SMPTE alignment tape on the machine. Observe the picture
at the demodulator output. Place the guide on manual operation (remove
automatic circuitry) , and operate the guide -position knob so as to move
the guide away from the head.
See Fig. 11 -22A. The highest- frequency burst packet is at 4.2 MHz. This
is sufficiently close to 4 MHz to make a point. If the guide is backed off so
that the first bar of band B is aligned with the fifth bar of band A (4 cycles
displacement) , the time displacement is 1 microsecond. (One cycle at 4
MHz is 1/ (4 X 106) = 0.25 µs, and since the displacement is four of these
cycles, the error is 1 µs.
With this same adjustment, observe the 2 -MHz burst. Since one cycle
at this frequency is 0.5 µs, the 1 -ps error will show alignment with the
third black line (2 cycles displacement) . There will be a slight difference
because the 4 -MHz reference is actually 4.2 MHz.
Now, if you were able to measure the actual radius of the pole -tip arc,
you would find it 1 mil too large. The rule of thumb is that 1 mil = 1 µs,
0.5 mil = 0.5 µs, etc.
Fig. 11 -22B shows the same amount of displacement, but playback is
on a headwheel with slightly less than 1 mil of tips remaining. Note that
the 4.2 -MHz sine waves are "breaking into black," indicating a loss of
signal at this frequency. A headwheel assembly is normally returned (ex-
changed) for a new assembly when the tips reach 1 mil.
Still another timing check you can demonstrate to yourself with the
SMPTE alignment tape is shown by Fig. 11 -23. Observe the demodulator
output on the monitoring oscilloscope while you adjust the guide farther
away from the head from zero alignment error, on the portion of the tape
containing the sin2 pulse.
Note that both the sync pulse and the sin2 pulse move to the right in a
faint trace. For example, assume you are using the Tektronix Type 529 cro.
The time base is set for 0.125H, and the 5 -times multiplier is used. Thus
the sweep rate is 0.025H /cm, or (0.025) (63.5) = 1.59 p.s /cm. So a 1 -µs
displacement is 1/1.59 = 0.62 cm (approx) This means that the faint
.
trace at the trailing edges of sync and the sin2 pulse will move to the right
by 0.62 cm when the guide is moved farther from the headwheel. This will
occur at the same time that the picture monitor shows the first cycle of
band B aligned with the third black line (second cycle) of band A on the
2 -MHz burst packet (Fig. 11 -22) .
Cycles 0 1 2 Cycles 0 2 4
Displacement Displacement
Band A
/
/
ll(l
/
First Cycle First Cycle

Band B Band B
r7/
/i////////
IE/
////////////
1/y,,,:
(i/ ,,,, 4.2 MHz
(A) Guide backed away 1 mil.
/////0///////il.
7/////://///////,.....
7Ì//I //I/fl////Í
, -
"Breakout" lirsl occurs at center

?MHz ofpicture (center of guide radiusi. 4.2 MHz
(B) Heavy breakout at 4.2 MHz.

Fig. 11 -22. Guide back -away method of estimating remaining
head -tip projection.
This particular procedure will serve as an excellent "warning flag" of

usable tip projection remaining on the head. For example, if the video head
is worn down to near the 1 -mil minimum projection, the 4.2 -MHz bursts
at the center of the picture will tear out with "whites" going to black.
This tells you that the head is starting to "break out" at this frequency, and
the head is nearing the end of its useful life.
Fig. 11 -23. Conversion of microseconds

of error to mils of tip projection.
Faint Traces
Many heads can be used down to about 0.8 mil before discarding, but
only at the risk of excessive clogging, noise, and lowered resolution. Also,
the recording- current drive (in the record mode) becomes rather critical
in "head optimization" for proper recording and playback ( next section) .
To avoid any confusion in your mind, answer the question, "what is
standard tip penetration ?" The tip penetration is equal to the tip projection.
This is to say that if the vacuum guide is adjusted to minimize skewing,
when playing back the SMPTE standard alignment tape, the penetration
of the video tips into the tape (tip penetration) will be equal to the video
tip projection. Therefore, the tip penetration is greater with a new head
than when this head is worn down. There is a greater rate of tip wear with
maximum tip penetration. At the same time, the greater penetration serves
as a "self- cleaning" agent for heads (minimum clogging from tape oxide
particles) and gives greater freedom from dropouts.
NOTE: The following information is included for a more complete under-

standing of tip projection but is not normally employed by the operator
in practice.
If the operator has a basic understanding of what constitutes standard

position of the vacuum guide, he can set this position without the standard
alignment tape, but not with the same laboratory precision. In practice,
the following information can be used to obtain a reasonably approximate
setting of the guide. (See Fig. 11-24.)
1. The distance from the axis of rotation of the heads to the center of
curvature of the vacuum guide is 1.0334 inches.
2. The nominal tape thickness is 1.4 mils.
3. The minimum radius of rotation of the magnetic -head pole tips is
1.0329 inches and the maximum is 1.0356 inches.
4. The vacuum guide is adjusted to the standard recording radius, which

is calculated as follows:
1.0334 inches (distance to guide)
-0.0014 inch (nominal tape thickness)
1.0320 inches
This figure (1.0320 inches) is necessary to find the "departure point,"
or the distance from the axis of rotation at which the pole tips just
touch the tape surface. This distance also defines the radius of the
pole -tip rotation.
1.0334"
For Zero Eccentricity
With Axis of Rotation)
Surface of Tape
Fig. 11 -24. Relationship of

Radius of Rotation
standard recording radius
to departure point. Axis of Rotation
li
i- Nominal Tape Thickness 1.4 Mils
5. Measure the diameter of the drum rim with a micrometer, and

divide this by 2 to obtain the radius.
6. Measure the actual tip projection above the drum rim with the tip -
projection gauge that is provided with the system.
7. Add the tip projection to the radius of the drum to obtain the radius
of pole -tip rotation.
8. With any prerecorded tape threaded, place the machine in the play
mode and back the guide away from the heads until the signal is just
lost, as revealed by zero rf output at the switcher.
9. Subtract the standard recording radius of 1.0320 inches from the
total pole -tip radius of rotation. Move the guide in from the depar-
ture point by this amount, as measured with the tip -projection gauge
jigged to the vacuum -guide block. This places the guide to the stan-
dard tip projection within the tolerances of measurement.
Two examples follow:

Example 1:
Drum Diameter = 2.064 inches
Radius = 1.032 inches
Tip Projection = 2 mils
Radius of rotation = 1.032 + 0.002
= 1.034 inches
Move in from departure point = 1.0340 1.0320 -
= 2 mils
Example 2:
Drum Diameter = 2.062 inches
Radius = 1.031 inches
Tip Projection = 2 mils
Radius of rotation = 1.031 + 0.002
= 1.033 inches
Move in from departure point = 1.033 - 1.032
=1mil
Note that in both cases the measured tip projection was 2 mils, but the
tolerances in the head drum resulted in a 1 -mil difference in the radius of
rotation. The foregoing procedure places the vacuum guide to the stan-
dard, but actual tip penetration depends on the effective radius of rotation,
and not on tip projection alone.
Fig. 11 -25 illustrates the foregoing procedure and gives the proper adjust-
ment from the departure point over the normal range of rotation radius.
NOTE: Some stations employ a dual standard positioning of the vacuum

guide. Tapes to be used only by the local station can be recorded at
lighter -than- standard tip pressures for new heads until the heads are worn
to the point where standard SMPTE projection can be used. This increases
tape life and results in less head wear per hour, if not carried to extremes.
A head assembly is normally retired (returned to manufacturer on an
exchange basis for a rebuilt assembly) at 1 -mil measured tip projection.
This is due to the fact that the pole -tip gap widens somewhat below this
projection and is apt to become useless during a recording or playback.
However, if too little recording radius is used, the tape may not play back
on a tip projection of, for example, less than 1.3 mils. This would require
premature retirement of the head assembly and be more costly in the long
run. The dual standard should not be used on an interchange basis, since
recordings are not completely interchangeable between heads at different
points on the head -wear curve.
Tip Projection
(Al Measure diameter with micrometer.
1131 Divide by 2 to obtain radius.
IC) Add tip projection above drum surface to radius.
ID) Back guide off to obtain departure point.
(El Subtract 1.032 inches from C.
IF) Move guide in from departure point to the value obtained in E.
3. i
3.
3.
3. 1
2. 3
2.
Tip Projection Center 2.25 Mils
2. 1
2. 0 1
1. 8 I
1. 6
1. 4 I
1. 2
1. 0 t
1 m11 1 mas 1 0339 1.0343 1.0347 1.0351 1.0355

1.0329 1.0333 1.0337 1.0341 1.0345 1.0349 1.0353 1.0357
M'n Radius of Rotation Max
Fig. 11 -25. Instructions for move -in from departure point.
Quadrature Adjustment
Electronic delay correction is employed in both the record and play
modes by RCA. Ampex supplies a delay equalizer as optional equipment
for the play mode only. In the following procedure, Steps -A, B, and C
apply to Ampex; Steps D through J are for RCA.
NOTE: On the latest Ampex and RCA systems, quadrature adjustment is
strictly a factory procedure in headwheel assembly. The description here-
with applies only to older systems, a number of which are still in use at
the time of this writing.
A. Reproduce the test pattern recorded on the alignment tape and

identify channel 1 by reducing the channel -1 gain control on the
switcher unit. Outputs appear in the picture in a 1, 4, 2, 3 order;

that is, channel 4 is directly below channel 1, channel 2 is directly
below channel 4, etc.
B. Observe the relative position of the four bands of information as
they appear horizontally on the picture, and choose a band which
occupies a mean position. Hereafter, use that band as a reference.
Assume here that channel 3 is the reference. With respect to channel
3, channels 1 and 2 are early; that is, they are displaced to the left
on the monitor. Channel 4 is late, displaced to the right on the
monitor.
C. The angular position of the Ampex 1000 -series head can be changed
by adjusting the tapered Allen-head screws. To advance a head, it
must be moved in the direction of rotation of the head drum; con-
versely, to retard a head, it must be moved in the direction opposite
to the direction of rotation. To advance a head (in our example
the channel -4 head) , first loosen the tapered screw leading the head,
and then tighten the tapered screw following the head. (This simply
shifts the quadrant which contains the head in the direction of rota-
tion of the head drum.) To retard a head (in our example channels
1 and 2) , first loosen the screw following the head and then tighten
the screw leading the head. Make these adjustments in small incre-
ments (not more than I/8 turn of the adjusting screws) checking
,
with the alignment tape after each adjustment. The head which
corresponds to a given channel can be identified by marks scribed
on the Bakelite hub of the slip -ring assembly just forward from the
terminals where the head leads are connected.
NOTE: Where the playback -delay amplifier unit is employed in the
Ampex system, playback adjustments may be made with this unit, and
the relative calibrations of the individual channel knobs may be read for
a direct indication of which segments need to be adjusted and the direc-
tion of adjustment. This unit also permits proper playback of any tape
made from another source which may have quadrature errors relative to
the Ampex head.
Steps D through J apply to RCA machines.
D. Place the guide -position switch on the control panel in the automatic
position.
E. Install the alignment tape, and place the tape recorder in the play
mode.
F. Adjust the control -track phase control to produce the optimum
picture. Set the four high- frequency compensation controls to zero.
G. Set all four delay knobs on the playback -delay amplifier to zero, and
observe the picture. The presence of steps in the vertical lines (Fig.
2 -19) indicates that the heads are not exactly in quadrature (90°
apart) Identify the head which produces an average amount of
.
horizontal displacement with respect to the other heads (steps about

halfway between the extreme left and the extreme right) , and leave
the delay knob corresponding to that head at zero. Adjust the other
three delay controls to minimize the horizontal displacements.
NOTE: Since the delay steps are quite small, this adjustment must be
made with care to insure optimum alignment of the vertical bars. In ad-
dition, penetration and scalloping adjustments should be retouched to
achieve the best results.
H. The quadrature relation of the heads is the same for record or play-
back, except that the relative sense ( lag or lead) is reversed. Con-
sequently, the delay controls on the record -delay amplifier should be
adjusted as follows:
(1) Note the channel on which the delay control is set to zero on
the playback -delay amplifier. Set the delay control for the same
channel to zero on the record -delay amplifier.
(2) Turn each of the other three delay knobs on the record -delay
amplifier to the same numeral as that of the corresponding play-
back delay knob, but in the opposite direction from zero. (For
example, if the delay knob on the playback-delay amplifier is
set to +3, set the delay knob on the record -delay amplifier to
-3.)
I. Remove the alignment tape from the machine.
J. Record a test pattern or other test signal with vertical lines; play-
back of this recorded signal should produce vertical lines. In addi-
tion, when the control -track phase knob is turned to another track
position, the vertical lines should contain a minimum of horizontal
displacements. Large displacements indicate that the corrections
should be rechecked.
NOTE: The video heads can be self- checked for quadrature quite simply,
but only after head optimization, as described in Section 11 -6, is followed.
The reason for this is that one head that may be overdriven with fm
signal in the record mode will play back with a slight effective quadrature.
The procedure is to make a recording of vertical lines and play back with
the tracking control adjusted sequentially through all four head -tracking
positions, while the operator watches the playback monitor. For example,
if head 1 recorded vertical sync, and tracking is adjusted so that the same
head plays back vertical sync, no quadrature error will be apparent even
though it may exist. However, if tracking is adjusted so that each of the
four heads picks up the track recorded by the adjacent head, quadrature
error between any two heads will be revealed.
11 -6. OPTIMIZING THE VIDEO HEADS

The record currents for each of the four video heads must be separately
adjusted (by means of record -channel gain controls) for optimum results.
Fig. 11 -26A illustrates the playback of a test pattern with proper deviation
and head optimization in the recording process. Fig. 11 -26B shows break-
out into black of white areas following black -to -white transitions; this can
be caused by overdeviation of the modulator (excessive level) with prop-
erly optimized heads, or by normal deviation but with one or more heads
overdriven with fm signal. (This effect can also be produced by improperly
adjusted playback -channel equalization controls.)
The result shown by Fig. 11 -26B is inevitably called overdeviation, even
though this is not necessarily the case. In fact, this photograph was made
from the playback of a tape with the test pattern recorded with exactly the
same deviation as that used for the tape of Fig. 11 -26A. The difference
is in the levels fed to the heads during the record mode. Since the optimum
record currents change with head wear (heads may wear unequally) , opti-
mization procedures must be carried out at frequent intervals. Wear is
greatest during the first hours of operation of a new head assembly; there-
fore, checks should be made approximately every 10 operating hours for
the first 30 hours, then after every 20 hours of use. Poor signal -to -noise
ratio or banding (bands or unequal color and contrast) usually indicate
the procedure is past due, regardless of hours.
When heads wear, there is less shunting effect of the pole tips, and less
signal is required for optimum recording. This is the reason why tapes
sometimes play back with what appears to be overdeviation, even though
normal video levels are maintained. Tapes which are very objectionable in
this effect can sometimes be greatly improved by removing all high -
frequency equalization (where used) in the playback channels. The result-
ing picture may be soft (lacking in fine detail) , but such softness is
normally less objectionable than the severe white -level noise. On a color
signal, this procedure cannot be carried to the point where there is in-
sufficient chroma and burst level for proper handling in the color processing
system.
Video -head optimization procedures have become well standardized
as follows:
(A) Properly adjusted,

[11 (B) Improperly adjusted.
Fig. 11 -26. Effect of adjustment of record -channel gain controls on playback
of tape recorded with test pattern.
1. Feed a signal with a large -area white reference ( such as a mono-

scope or window signal) to the system. Check to see that the devia-
tion is normal.
2. Place the system in the record mode. Speak into the microphone,
and identify channel 1 by turning the channel -1 record gain control
fully counterclockwise (to zero) . For example: "This is channel 1
on zero."
3. Rotate the gain control clockwise, stopping momentarily on each
dial number to identify this number aurally. When the maximum
clockwise position is reached, reverse the gain, and again identify
each dial number at the momentary stops. When finished, return
the gain control to midpoint.
4. Carry out the same procedure for channels 2, 3, and 4. Rewind the
tape to the start of the recording.
5. With the scope set on the wideband -response position and con-
nected to the switcher -unit output, play back the recording. Remove
any high- frequency compensation that may exist in the playback
circuits. Keep the playback -channel gains and scope gains low
enough that there is no chance of rf compression.
6. Track the playback so that the same head which records vertical
sync is playing back vertical sync. Watch the scope display of the
switcher output. Monitor the voice identification of the channel-1
gain. At some point, the rf envelope of channel 1 will cease to in-
crease as the gain is raised. Make a record of this point. A further
increase in record current causes the playback level to decrease.
7. Repeat Step 6 for each of the remaining channels.
8. Set the record gain control for each channel to the point at which
the gain for that channel was maximum. Make a new recording.
Rotate the head record -current meter switch to each channel in turn,
and make note of the current readings for reference. When this
recording is played back, should any banding be apparent, identify
the head (s) causing the banding, and repeat the procedure of
optimization for these channels. Be sure all playback-channel gains
are equal, as shown by Fig. 11 -27A. (Ignore the vertical -sync tip in
setting gains.)
NOTE: With regard to the rf patterns of Figs. 11 -27A and 11 -27B, the
level controls should be adjusted for equal levels as in Fig. 11 -27A if the
pattern of Fig. 11 -27B is observed. For the pattern of Fig. 11 -27C, the
point of observation is before agc action, and no manual gain controls
are provided. For optimization, the individual head record currents are
still adjusted for maximum level of each individual head on playback.
Actually, the record current should be increased just slightly past the
point of reaching maximum playback level. (In Fig. 11 -27C, the rf enve-
lope is chopped to ground reference to indicate each individual channel.)
yent -Sync Tip
(A) Equal gains. (B) Unequal gains.
(C) TR -70 system.
Fig. 11 -27. Switcher- output rf waveforms.
9. Fig. 11 -28 shows the oscilloscope and monitor displays if the

switcher phasing is improperly adjusted ( switcher phasing on
Ampex, 960 delay on RCA). Adjust for the elimination of holes
on the scope display and white streaks in the picture. The white
(A) Switcher output waveform. (B) Monitor presentation.

Fig. 11 -28. Switching pulse misphased.
streaks are similar to dropouts, except that switcher phasing error

is synchronous, whereas dropouts are random.
10. It should be common practice during any of the preceding record
modes to check the phasing of the control track relative to the edit
(hence reference) pulses. Proper adjustment is shown in Fig. 11 -29.
If two tapes are spliced together with slightly different control -track
phase relative to the recorded tracks, the effect in Fig. 11 -30 will
occur. If the phase difference is large, complete loss of signal may
result, as the heads are between tracks. It is then necessary to
readjust the tracking control manually immediately following a
Fig. 11 -29. Waveform- monitor display Fig. 11 -30. Monitor display of poor
showing proper phasing of control- signal -to -noise ratio caused by
track signal with edit pulse. slight mistracking.
splice. (On the RCA TR-70 system, the monitoring cro presentation
of the record control track shows the edit pulse inverted, but proper
polarity actually exists at the head.)
The operator is now ready to make a short trial recording as a double

check on system operations. All of the preceding procedures, of course,
are not necessary before each recording. The entire setup procedure has
been described in the interest of completeness. Experience with a particular
installation will dictate the setup scheduling for optimum results on a
consistent basis.
11 -7. CHECKS DURING RECORDING

"l he following checks should be made periodically during a recording:
1. Control -track meter (or cro presentation) for normal current
2. Master- oscillator meter (or panel light) for proper erase current
3. Record -current meter for proper head currents as called for in the
optimization procedure
4. Head -servo waveform for stability
5. Capstan -servo waveform for stability
6. Audio -head erase meter for proper erase current
7. Audio -recording level on VU meter

8. All warning lights for any indication of malfunction
11 -8. CHECKS BEFORE PLAYBACK
The following procedure should be followed before playback of a tape:

1. If the recording was made on the same head used for playback, adjust
the control -track phase ( tracking) control so that the same head that
plays the vertical sync also records the vertical sync (head 1 for RCA
and head 4 for Ampex) . This procedure will result in optimum match
of heads. However, if the recording was made with a different head,
the capstan tracking should be adjusted to obtain maximum rf output
from the switcher unit. This normally indicates the best match of
heads.
NOTE: Where Intersync or Switchlock is used, a given head must play

back vertical sync for proper operation, depending on the particular
system and mode of operation.
2. For playback of any tape, the guide servo is normally placed in auto-
matic control so that skew is eliminated. If the automatic circuitry
should fail to work and time is not available to repair it, simply place
the servo in manual operation and adjust the position (tip-projection)
control to eliminate skew at the demodulator output, so that the
burden on the ATC system is minimized.
NOTE: It is important to bear in mind that the mechanical adjustment

should not be changed from standard. The worst condition that can exist
is that scalloping can be so noticeable in the picture ( which may be a
recording made elsewhere) as to require guide- height adjustment, which
must be done mechanically. If the mechanical adjustment for skew has not
been tampered with, it is normally a simple matter to give the height
adjustment a slight turn to eliminate the scallop. When this is necessary,
remember to recheck the height adjustment with the standard alignment
tape before making a recording.
3. Adjust the playback -channel gains for equal outputs, as shown in

Fig. 11 -27A (rf indication, and where gain controls are provided) .
4. Adjust the individual -channel high- frequency compensation controls

(where used) for the "snappiest" picture without transients, and for
best match of individual head response for contrast (minimum band-
ing).
5. Adjust the processing-amplifier controls for proper video, setup, and
sync levels to the line output.
6. Set the audio -output level control for standard level.
7. Rewind the tape and cue in to the proper roll point.
11 -9. EMERGENCY OPERATIONS
We will now take up color -tape problems in general -what to do when

you have trouble getting even a semblance of "lock-up" on a color tape for
purposes of setting it up for proper playback.
The color mode of operation demands the tightest tolerance in video and
servo functions. If the system will not find stable lock, temporarily forget
color! Go back to the "loosest" servo mode, tonewheel for RCA, or normal
for Ampex. If the system has switchable time -base error correction, turn it
off (bypass). Reduce all playback equalization to a low level. Sometimes
high- frequency transients from accumulated errors will prevent servo lock-
up. You can always bring these back to normal after lock-up has been
achieved.
The idea is to remove all feedback information from the demodulated
tape signal into the servos. This enables you to minimize all errors of the
head itself to match the tape to be played back.
Remember that time -base error circuitry is stable only over a limited
range. If accumulated errors exceed about 1 µs, the resulting variable -
delay modulation exceeds the practical range of Z. of the voltage -variable
delay line, "kicking" the servos and preventing lock -up.
As soon as you have obtained stable lock in the simplest form of servo
mode, you must make the head match the tape as closely as possible
(scallop and skew) . You observe the demodulator output on the picture
monitor to do this. Include quadrature adjustments if these are provided.
Then activate the ATC or Amtec ( while still in the nonsynchronous
servo mode, no tape -signal feedback) , and observe the error signal remain-
ing on the cro at the ATC or TEC error monitoring point. After consider-
able practice, you can actually go through all the head adjustments to obtain
minimum error signal on the cro at this point.
If the calibration is correct, the time -error correction signal will show
100 IEEE units (0.7 volt) peak-to -peak for a 1 -µs error. If this is ex-
ceeded, chances are that you will not be able to use synchronous servo
modes of operation; hence there will be no color lock. It is still better to
play the program in monochrome than to lose the show entirely.
For a color tape to play properly, the accumulated errors ( including
recording jitter) must be less than 1 µs. Select the ATC or TEC position
of the monitoring scope. On the latest systems, the error waveform may
appear as in Fig. 11 -31A. The "line" of errors may bounce up and down.
If the range of bounce is greater than 100 IEEE units, the trouble may
have been in the initial recording.
On older systems, you may obtain the waveform of Fig. 11 -31B. This
is a "locked error" simply showing the excursion of the time -error correct-
ing signals. Again, this must be under 1 µs, or 100 IEEE units.
In the case of a "random" walk of error signals as in Fig. 11 -31C, the
waveform will "float" at an unlocked rate. (When the cro selector switch
is in this position, the cro is externally synchronized to a 240 -, 480 -, or

960 -Hz sweep) .
Let us consider an example. If you have an early Ampex VR- 1000A, or

an RCA lA without time -error correction circuitry, make a recording on
this machine. Then make a dub of this on your latest system designed for
color, with the machine which made the recording supplying the dub signal
source. Chances are, in playback of this dub you will obtain the waveform
of Fig. 11 -31C, and you cannot go to a synchronous mode of servo opera-
tion such as Pixlock (automatic) or Linelock (horizontal) Early servos .
were quite "loose," particularly without time -base error video correction
circuitry.
loo
r
Bounce
o 0
(A) Waveform with bounce. (B) Locked error.
110
Free -Run
(C) Free -run error.
Error
Fig. 11 -31. ATC or TEC error waveforms.
There is a way you can check your system to differentiate between play-
back stability and jitter in the initial recording. Think of it this way: If
you place the picture monitor on external sync (local sync) and play back
the tape in a nonsynchronous mode ( tonewheel or normal) , then you
observe the total jitter of recording and playback. Maintaining this same
type of operation, superimpose the tonewheel or stability dots onto the
picture monitor. These are pulses from the head tachometer; therefore the
motion of these dots relative to the picture itself is an indication of play-
back jitter only. Now put the picture monitor on internal -sync operation.
This means the sync from the video -tape playback is separated and used
to sweep the monitor. Therefore the playback jitter is taken up automati-
cally in the monitor, and the tonewheel or stability dots will now indicate
recorded jitter only. It makes no difference whether you are operating the
tape system in a nonsynchronous or synchronous mode.
Caution: Some systems automatically switch the picture monitor to
external sync in going to the synchronous mode of operation, so it is best
to leave the tape system in the tonewheel or normal mode if you are not
sure. But check this for your particular system as soon as possible so that
you know your system.
The tachometer dots superimposed on the picture should never be so
"rapid" as to cause a "double- dots" appearance. If this shows up, either in
playback jitter or record stability, the chances are that you will not obtain
synchronous operation. A wide excursion (several inches), if at a slow
rate, is not an indication of jitter.
Obviously if this jitter shows up only as a recorded jitter, you cannot do
much to save the program, except to play it in a nonsynchronous (mono-
chrome) mode. If jitter shows up in the playback check, the servos need
attention. There is no substitute for the manufacturer's instruction book
here, but Chapter 12 will help give you a better understanding of servo
setup and troubleshooting.
After you have matched the head as closely as possible to the tape in
nonsynchronous operation (minimized errors as much as possible), you
can then try synchronous operation (with ATC or TEC on) If you can.
now achieve lock -up, readjust the individual channel equalizers and go on
from there.
The important thing to know is what was just described, how to tell
whether a particular tape is suitable for synchronous operation. You also
need to be able to tell whether the servos are at fault (playback jitter) or
the fault is in the particular tape recording involved.
Some erratic conditions, such as video tear -out on a certain head or heads,
can often be blamed on improper head antiresonance adjustments in the
latest systems where this feature is included. This adjustment (or adjust-
ments) compensates for the resonance of the head during playback and is
made by operating reactance (usually a capacitor) and resistance controls
on each channel playback amplifier while observing a sweep pattern (in-
jected by a special probe on the headwheel) at the fm switcher output. For
example, the procedure for the RCA TR -70 is as follows:
1. Place the machine in the stop mode, and remove the tape from the
headwheel panel and audio heads (rewinding the tape is not neces-
sary) .
2. On the switch panel just above the modules, set the TEST SELECT
switch to Res (resonance) and the CHANNEL SELECT switch to 1.
3. Remove the head resonance probe from the HD RES PROBE socket on
the switch panel above the modules by rotating the probe clockwise
approximately 45° and pulling the probe out. (Later RCA head -
wheels have a built -in probe.)
4. Insert the extruded portion of the head resonance probe into the
slot provided for it on the headwheel panel, so that the U- shaped
wire loop is close to the headwheel.
NOTE: The HD RES PROBE socket contains an interlock switch that pre-
vents the machine from being placed in any operating mode when the
head resonance probe is removed from the socket.
5. Push the TEST switch /indicator on the switch panel above the
modules.
6. Press the FM LEVEL button on the cro switcher panel.
Typical Ideal Waveform 1 Typical Ideal Waveform 2
Reactance and Resistance Controls Set Correctly Reactance and Resistance Controls Set Correctly
Reactance and Resistance Controls Set at Minimum Reactance and Resistance Controls Set at Maximum
Fig. 11 -32. Head antiresonance adjustment.
7. Manually rotate the tonewheel so that the notch on the tonewheel

is lined up opposite the tonewheel head. Observe the cro presenta-
tion of head 1, and slowly rotate the tonewheel for maximum ampli-
tude on the cro presentation.
8. Alternately rotate the RESISTANCE and REACTANCE controls on the
playback amplifier 1 module in either direction to obtain a smooth
or linear section on the waveform just prior to its rapid drop in
amplitude (Fig. 11-32 ) .
NOTE: The smooth or linear portion of the waveform may slope in either
direction. Proper adjustment results in the desired smooth portion regard-
less of direction of slope.
9. Move the CHANNEL SELECT switch to position 2.

10. Manually rotate the tonewheel one -quarter turn in a clockwise direc-
tion (looking from the tonewheel end of the headwheel panel) so
that the tonewheel notch is located at the top of the headwheel
motor, which is 90° from the tonewheel head. Observe the wave-
form presentation on the cro, and position the tonewheel for maxi-
mum amplitude from head 2.
11. Repeat the procedure described in Step 8, except that now the pro-
cedure is for the antiresonance adjustment for head 2 on playback
amplifier 2.
13. Manually rotate the tonewheel one -quarter turn in a clockwise direc-
tion so that the tonewheel notch is located 180° from the tonewheel
head. Observe the cro, and position the tonewheel for maximum
amplitude from head 3.
14. Repeat the procedure described in Step 8, except that now the
procedure is for the antiresonance adjustment for head 3 on play-
back amplifier 3.
16. Manually rotate the tonewheel one-quarter turn in a clockwise
direction so that the tonewheel notch is located 270° from the tone -
wheel head. Observe the cro, and position the tonewheel for maxi-
mum amplitude from head 4.
17. Repeat the procedure described in Step 8, exceiit that now the pro-
cedure is for the antiresonance adjustment for head 4 on playback
amplifier 4.
18. Press the STOP switch on the record or play panel to release the test
switch -holding circuit.
19. Replace the head resonance probe by inserting the probe into the
HD RES PROBE socket and rotating the probe 45° in a counterclock-
wise direction.
Erratic tracking problems are sometimes due to insufficient control -track

level on the recorded tape. The peak -to -peak level of the control -track
waveform is not an indication of either proper or improper control -track
level. Note from Fig. 11 -33 that the proper control-track level is indicated
by a slight shoulder at the zero axis of the distorted sine wave. This
amplitude is originally set as follows:
1. With the machine in the master record mode, the control -track play-
back button on the monitoring cro is depressed to observe the
simultaneous control -track playback presentation during the record

mode.
2. The control -track record -level control is then adjusted until the
simultaneous control -track signal just begins to show saturation, as
in Fig. 11 -33B.
3. While still recording, change the cro monitoring selector push button
from control -track playback to control -track record. Adjust the
frame -pulse amplitude to 1.5 times the sine -wave amplitude, as in
Fig. 11 -34 (waveform A) . Note that with this amplitude, it appears
that the frame pulse is not exactly at the peak of the sine -wave. To
properly judge this, however, it is necessary to reduce (temporarily)
the frame -pulse amplitude and adjust the frame -pulse phase control
for exact centering, as shown by waveform B of Fig. 11 -34. Then
when the pulse is readjusted for proper amplitude, waveform A
results.
NOTE: It was previously mentioned that the frame pulse is a negative -

going pulse starting from the crest of the current waveform. It was also
mentioned at that time that the RCA TR -70 monitoring position shows
a polarity the reverse of what actually exists at the control-track head.
For this reason, the frame pulse is shown starting from the valley rather
than the crest of the waveform, and it is positive-going. This is immaterial
to proper adjustment but is important to understanding the subject.
i
Shoulder
(A) Current too low. (B) Proper current.
rv\
J (C) Current excessive.
Flat Shoulder
Fig. 11 -33. Simulplay waveforms for adjustment of control -track

record current to proper level.
15
Correct
Frame-Pulse 10
Amplitude
Frame Pulse
To setframe -pulse
phase correctly,
reduce amplitude
to very low value.
O
Fig. 11 -34. Control -track record waveforms for RCA TR -70 recorder.
11 -10. ELECTRONIC SPLICING
The following basic outline serves as a guide for preparing to do an

electronic splicing job:
1. Determine what is to be accomplished.

2. Determine the relative timings. (Where are they to occur? )
3. You might desire to utilize the cue track for splicing information.
(The splice must be initiated 1/2 second ahead of the actual cut. This
is because there is 1/2 second, or about 15 frames, between the master-
erase and video heads.)
4. Should the audio track be retained?
5. Make certain that the machine is operating properly and that the
tapes used have edit ( frame) pulses on the control track. Without
these pulses, the electronic splicer will not turn on.
6. Make a few practice splices on spare tape to become familiar with the
technique.
7. If a second tape recorder is used as a signal source, it must be adjusted
to appear as a line feed as follows:
A. Servo mode in Pixlock.
B. Burst phase adjusted to give proper flesh tones.
C. Horizontal phase (in the splicer) adjusted to give exact leading -
edge coincidence at the input to the splice recorder.
D. If inserts or wipes are anticipated, set the system burst phase of the
feed machine correctly to match the burst phases of other sources
at the switcher-system input.
Following is a brief outline of the RCA electronic splicing technique.

This information is included here to acquaint the reader with the opera-
tions involved.
1. Put the machine on which the splice is to be made in the Switchlock

mode.
2. Select the proper splice mode. (Insert retains the old control track;
Add -On puts down a new control track.)
3. Select the proper audio mode (audio off to retain old audio) .
4. On the monitor selector switch, press both VID IN and DEMOD buttons.
(On the TR -70, use PULSE CROSS.) With the switch in the E -E mode,
note the normal phasing horizontally between the superimposed
signals in the sync region (Fig. 11 -35) .
5. Play back the tape on which splicing is to be done. Punch the ATC
button on the cro and adjust the head for minimum errors.
6. For the add -on mode only, perform capstan adjustment as follows:
Punch the CAPSTAN button on the cro. With the tape playing, put
the switch on the splice logic module in the Capstan Adj position,
and adjust CAPSTAN ADD ON ADJ (same module) for minimum shift
of the sampling pip on the slope. (Switch between Capstan Adj and
Fixed TW and adjust until the shift of the pip is zero.) Be sure to
return this switch to Var TW.
7. While still playing back the tape, adjust SPLICE RECORD H PHASE (on
splice timing module) until the phasing on the monitor (looking at
VID IN and DEMOD OUT superimposed) is the same in playback as in
the E -E mode (Step 4). In Fig. 11 -35, the delayed signal is the
demodulator output. To set the splice record horizontal phase control,
adjust this control so that the playback gives no vertical displace-
ment and lias the approximate horizontal position shown. (The
heavy, or darkest, portion of the hammerhead is the overlap interval
of the two signals.)
NoVertical Line
Displacement
Fig. 11 -35. Simultaneous presentation of video input and demodulator output

with pulse -cross display on picture monitor.
8. Go into regular splice operation. The REC button must be pressed 1/2
second before the splice is to be made. Then the PLAY button must
be pressed 1/2 second before the end of the splice. Remember that
you must go into the play mode before the record mode; otherwise,
the machine will make a regular recording, ignoring the splice mode.
NOTE: This section has discussed only the basic technique of electronic
splicing. The use of more refined automatic editing programmers varies
considerably with every make and is beyond the scope of this book. The
operator familiar with Chapter 10 and this chapter will be able to master
those techniques more quickly by following the operational outlines ac-
companying the equipment.
EXERCISES
Q11 -1. Where should the master equalizer control be set?

Q11 -2. Why was the setting of individual playback levels not mentioned?
Q11 -3. If the individual channel equalizers are adjusted for even "tops"
across all heads as observed at the demodulator outputs, and the
"bottoms" are unequal, what does this mean?
Q11 -4. What is the proper way to adjust monochrome and color time -base-
error gains?
Q11 -5. When the demodulator gain is adjusted for proper output in the E -E
circuit mode, should this level be the same for any tape playback?
Q11 -6. Assume the demodulator balance control is set in the E -E mode, but
on playback of a tape, it is necessary to readjust this balance for
minimum rf in the blanking and sync regions. Is this normal? (Usu-
ally occurs only on older low -band systems.)
Q11 -7. When you are checking signal inputs on an external scope without
the IRE (IEEE) graticule, what is the proper voltage proportion for
video and sync?
Q11 -8. If you have a 0.2 -/..ts ringing period of a sine- squared pulse, what is
the cutoff frequency?
Q11-9. What are the specific K- factor characteristics to look for in the (A)
low- frequency, (B) high- frequency, and (C) midfrequency region?
Q11-10. When it is necessary to use special -effects mixing on a color tape
playback, which should be adjusted first, system phase or burst phase?
Q11 -11. If you make a recording of the sine- squared pulse- window test
signal for checking system performance, how do you differentiate
between recording and playback characteristics?
Q11 -12. What is the gamut of rf frequencies covered in modern high -band
tape-recording systems?
Q11 -13. In either the RCA or Ampex high -band tape systems, the video
processing amplifier can be operated in one of two modes, the
external- reference mode or the regenerated -sync mode. What is the
difference?
Q11 -14. Does a "standard test tape" serve as a standard for both new and
worn heads?
Q11 -15. What is the minimum tip projection normally allowed on heads
before they are returned to the factory for exchange?
Q11 -16. Is it advisable to reduce tip penetration below the "standard" to ob-
tain greater life of the headwheel?
Q11 -17. How do you lap- dissolve between a studio camera and a tape
machine or between two tape machines?
Q11 -18. What can be wrong if no movement of the tape occurs in a play
mode but everything is actually working properly on the machine?
CHAPTER 12
Quadruplex Tape System

Maintenance
This chapter concerns general testing and maintenance techniques as

applied to any quadruplex recording system. Lubrication schedules recom-
mended by the manufacturer for the specific system used should be faith-
fully executed. This is of extreme importance due to the electromechanical
nature of the tv tape system. Complete service records of work performed
or adjustments made, correlated with the elapsed-hour meter, are invalu-
able for increasing efficiency in tests and maintenance.
12 -1. OVERALL PERFORMANCE EVALUATION

An overall performance evaluation of a television -tape system helps to
pinpoint the need for maintenance in specific areas. A professional job of
cleaning the tape transport and video -head assembly is of vital importance
before testing or maintenance procedures are conducted.
Major troubles that can result from lack of cleanliness may be sum-
marized as follows:
1. Dropouts (white flashes) due to oxide accumulation between heads.
2. Dots at a 960 -Hz rate as a result of scratches on the coated surface
of the tape caused by oxide accumulation.
3. Oxide in the vacuum guide or hose, causing scallop or skew due to
lack of concentricity with the tape. A wandering type of skewing can
be caused when oxide which sheds into the vacuum guide alternately
clogs and clears the air path. This effect can also be caused by an
alignment tape with many hours of use-or even after only a few
hours if the transport tensions are such as to cause excessively rapid
stops or jerking of the tape when the STOP button is depressed. An-
other cause is bent reels which place undue stress on the edges of the
tape.
311
4. Contamination of the nylon bearings under the vacuum -guide block

or a contaminated screwhead contact with the arm that is actuated
by the guide servo (to position the vacuum guide) This can result
.
in a slightly different tip penetration each time the tape is stopped

and restarted.
5. Erratic tracking caused by tape slippage resulting from a contaminated
capstan or capstan pinch roller.
CAUTION: New tape from the factory will often exhibit what seems
to be an excessive number of white flashes (dropouts) when the first re-
cording is played back. When tape is first received, a spot check should be
made (recording sync only for about the first 5 minutes, 5 minutes near
the middle, and 5 minutes near the end of the reel) to observe the playback
raster. If dropouts are numerous, the tape should be polished by allowing
the entire reel to run through the heads two or three times. A worn head
is best for this polishing operation. This is not quite so important when a
dropout compensator (Chapter 8) is employed.
Overall evaluation can be an evaluation of the recording and playback

functions on a routine operational basis considering only the composite
video signal, or it can be an evaluation which also includes the following
items:
1. Time required after tape roll time for complete stabilization of the
servos.
2. Maximum input level to the video -processing amplifier tolerable
before clipping occurs, and minimum input level tolerable before sync
gating and /or sync formation is affected. This is a means of antici-
pating trouble before the unit ceases to function properly over the
normal operating range.
NOTE: The trademark of a qualified maintenance man is knowing his
safety margins. When the safety margin begins to slip, he goes to work.
In fact, most of the effectiveness of any preventive- maintenance schedule
is in knowing and using the safety margin.
3. Evaluation of adjustment control ranges. Some controls are normally

set at one extreme (cw or ccw) Others, however, such as frequency
.
controls, pulse- widths controls, and video -level controls ( including

blanking and sync) should be checked to determine if proper opera-
tion is obtained near an extreme end of the range.
4. Modulator-demodulator amplitude linearity range, the extent of over-
deviation possible before amplitude compression occurs or the pic-
ture plays back with visible transients in black -white transitions.
NOTE: The video heads must first be checked for optimization
(proper driving currents) before this evaluation (described in
Chapter 11) .
5. The playback signal -to -noise ratio of a recording made by the system
concerned. This is an excellent indicator of any system degradation
other than amplitude- frequency response. ( Amplitude- frequency re-
sponse is usually revealed by evaluation of the composite video signal.
The maintenance engineer should bear in mind that any standard test
tape such as the one described in Chapter 11 is a check of the playback
function only. For overall record -playback checks, it is necessary to have
access to the necessary test -signal generators. Conventional signal -analysis
techniques are used when dealing with individual test signals. The test -
signal input to the recorder must be a standard composite signal. Some
older test -signal generators contain only horizontal sync so that clamping
circuits will function properly. Fig. 12 -1A is the vertical -rate display of
the output of a typical older stairstep generator; this display indicates that
horizontal pulses are contained in the vertical -sync interval. A television
tape recorder will not make a satisfactory recording of this signal. The same
(A) Direct output, showing (B) Output of unit which inserts

horizontal -rate pulses in vertical sync. keyed blanking and sync pulses.
Fig. 12-1. Vertical -rate display of typical stairstep- generator signal.
signal fed through a unit ( such as a monoscope amplifier with external

input provision) with keyed blanking so that horizontal pulses are elimi-
nated during the vertical interval is shown in Fig. 12 -1B. This type of
signal is necessary to make a recording on television-tape systems which
generate a reference pulse at a specific time in the vertical- synchronizing
interval ( all recorders after the Ampex 1000) The vertical pulses are not
.
only integrated, but they are also effectively differentiated so that a par-
ticular time reference is available for generation of the reference pulse.
When a signal such as that in Fig. 12 -1A is fed to the system in the record
mode, no reference pulse is generated (or if it is generated, it will be
erratically keyed) , and playback may be either erratic or impossible.
NOTE: See Section 12 -4 for special test procedures employing noncom -
posite signals.
39 -ï TELEVISION BROADCASTING
12 -2. THE VIDEO -HEAD ASSEMBLY
Maintenance of the video -head assembly involves primarily cleaning,

proper positioning of the vacuum guide, and video -head optimization.
All of these items have been covered previously.
If the head clogs easily, measure the tip projection as described in
Chapter 11. Heads should normally be retired when they become worn
down to a 1 -mil tip projection, as a slight widening of the gap is apt to
occur at any point below this value. The gap is quite easily clogged by
iron -oxide particles from the tape and may become useless on the passage
of a splice. NOTE: If a rolling- through amplitude change occurs during a
program as observed on the rf envelope from the switcher (with resultant
degraded and banded picture) , try holding a very soft brush (moistened
with Freon TF) lightly against the rotating pole tips while they are in
motion. Some operators grasp the vacuum guide gently with the thumb
and middle finger to steady the hand while extending the forefinger (in
the direction of head rotation) and lightly grazing the pole tips as they
rotate. While this method must be used with extreme care to avoid injury
to the finger, a clogged head is almost immediately cleared.
There are two head faults over which the operator has no control:
azimuth alignment and axial position (see Figs. 2 -20 and 2 -21) However,
.
it is possible to check the heads for these conditions; such checks should
normally be made by the maintenance engineer immediately on receipt
of a rebuilt head assembly.
Check for Azimuth
A head with gap tilt lays down a narrow track, as shown in Fig. 12 -2A.
This results in a loss in amplitude, bandwidth, and signal -to -noise ratio.
An exaggerated azimuth error is shown in Fig. 12 -2B to illustrate how the
recorded track is narrowed. This is equivalent to widening the playback
gap of a normal head. The tilt of tracks recorded when the gap is not
parallel to tape travel is shown in Fig. 12 -2C, and the method used to
detect this error is given in Fig. 12 -2D. Make a recording of vertical lines.
Playback should be done with the tracking adjusted so that the same head
that plays vertical sync also recorded vertical sync. When tracking is per-
fect, the head gaps are directly over the recorded tracks. By slightly mis-
tracking on each side, any apparent quadrature effect will reveal azimuth
error. Mistracking in the direction of 1 (Fig. 12 -2D) will retard the band
in time, and mistracking in the direction of 2 will advance the band in time.
If the lines are parallel to tape travel, no error will occur. The head assem-
bly must first be optimized and aligned properly, as outlined in Chapter 11.
Check for Axial Position

As shown by Fig. 2 -21, axial misalignment of a head results in non-
standard spacing for one band of 16 lines. After the heads have been
Parallel to
Tape Travel
Rotation
1 Gap
- - 10 Mils
Tape
Narrow Track Width Normal Track Width I l
I I
II ll
I l
(A) Wheel. (B) Exaggerated tilt of gap.
Time Difference
_1
Head Gap
¡Rotation
1deoTracK
Head
Gal
Azimuth Tilt -WI Normal

Azimuth Error
(Paraliell Retarded in Time
Head Mistracked in One Direction
rTape 01 i
Head Mistracked in Other Direction
I
ÌJ
Advanced in Time
(C) Track tilt. (D) Mistracking.

Fig. 12 -2. Azimuth alignment check.
optimized, make a recording of a suitable signal (a test pattern is best) and

play the tape back with the tracking control adjusted so the head that
records vertical sync plays it back. The playback picture should be optimum
under this condition. Now change the tracking by one head. (This is done
simply by moving the tracking control either clockwise or counterclockwise
until lock -in is again achieved.) If axial misalignment exists, one band
will become noisier than the rest. This will result for three out of four
head trackings. In many cases, an effect known as half-banding occurs; this
simply means that approximately one-half of a given band is noisy. Half-
banding results when the pole tip scans properly across a portion of the
track, but the remaining portion of the track is curved relative to the
standard track. Thus, a uniformly high noise level throughout one band
indicates that one head is not in the same vertical plane (axial position) as
the other three heads. Half- banding results when the plane of pole -tip
rotation is not perpendicular to the tape. In either case, the manufacturer
will allow full rebate on such a head assembly if it is returned within the
first 10 hours of operation. (Check your manufacturer on this, as this "free
time" may be lengthened.)
Tip Ringing
In color, if three or four bands of lines (indicating one of the four video
heads) show excessive moiré, this may be caused by tip ringing. Assuming
the playback level and equalization for the particular head in question
have been properly adjusted, tip ringing can cause a color interference
pattern (moiré) even when not very noticeable in monochrome.
Tip ringing is a head pole -tip fault caused by deterioration of the
magnetic -gap material. The tip can actually dig into the tape in a severe
instance of this condition and cause excessive tape shredding with possi-
ble tape damage.
12 -3. OPTIMIZING THE LOCAL SYNC GENERATOR

Requirements for certain characteristics of the local sync generator are
more exacting for integration with tv tape systems than for other studio
equipment. Optimizing the station sync generator indicates neither a differ-
ent adjustment nor a nonstandard adjustment to make it compatible with
the tape recorder. It does, however, indicate more critical adjustment of
frequency and more thorough maintenance of stability of counters, afc
circuits, and pulse distribution. This is particularly true of older sync gen-
erators and tape systems.
After the sync generator has been adjusted according to the manufac-
turer's instructions, perform the following check:
Set the master oscillator in the free -running mode. Insert the scope probe
to observe any 60 -Hz signal (such as counter -chain output, vertical drive,
or blanking) with the scope trigger selector on the 60 -Hz line position.
This reveals any slip of the vertical- frequency generator output with respect
to the line frequency. Adjust the master -oscillator frequency so that the
trace is stabilized with the line frequency; this provides a fine adjustment
of the oscillator frequency, provided that the counters are properly func-
tioning and centered. A very slight drift back and forth may occur, but no
sudden changes should exist.
Setting the master oscillator control to line -lock should immediately
stabilize the trace after the initial phase slip to lock. If this does not occur
or if the trace becomes unstable on the line -lock position, the afc circuitry
is in need of service. When individual counter stages employ adjustable

controls, always check to see that such controls are centered in the range
midway between the extremes where proper countdown is lost.
With the sync -generator frequency adjusted, no difficulty should be ex-
perienced in tape-system operation using local sync as the playback refer-
ence. The same is true when the Ampex Intersync or RCA Pixlock mode
is employed with the local sync generator in the Linelock position. How-
ever, both Ampex and RCA recommend using crystal control for the local
sync generator in the latter case for maximum playback stability. When,
for any reason, it is desirable to operate in the Linelock sync position, the
preceding fine tuning of the master oscillator along with stable afc circuitry
will normally result in satisfactory operation. Modern systems, of course,
employ "color lock" only.
Sync cross talk is another problem. This term applies either to cross talk
within the sync generator itself or to the "windshield -wiper effect," similar
to the horizontal motion resulting from cochannel interference on a home
receiver.
Cross talk within the generator itself is usually caused by very small
leakage of any of the counter frequencies to the master oscillator. This
trouble is most evident on video monitors employing pulse-width (Syn-
chroguide) horizontal circuitry. Cross talk produces a slight horizontal -line
displacement at the vertical raster edges and vertical lines in the picture.
The engineer can check by observing any pattern consisting of vertical lines
(such as a grating signal or keyed -burst signal driven from local sync) ,
preferably on a Synchroguide-type monitor. The frequency of any existing
cross talk can be determined by considering the number of horizontal line
displacements occurring from top to bottom of the raster, as follows:
Number of
Frequency Displacements
4500 Hz 70
900 Hz 14
180 Hz 3
Thus, cross talk from the 900 -Hz output of one of the counters results in
14 displacements, which is quite close to the 16 -band ( 960 Hz) raster con-
struction of a tape-recorder output. Sometimes a scallop effect, which can
cause the operator to misadjust the vacuum -guide height adjustment in an
attempt to eliminate an error, is actually caused by the local sync generator.
However, this effect will be observed at the input to the tape system when
cross talk exists. Certain types of master monitors employ a synchronization
which, for all practical purposes, results in line -to -line sync. This tends to
start each active line of picture information at the same spot following
horizontal sync and minimizes any effect of line displacement that normally
would be quite apparent on an average -sync monitor (and on most modern
home receivers). For this reason, the station should monitor tv -tape systems
with an averaging -sync monitor.
When "synchronous- type" cross talk is found, the indicated counter
should be additionally shielded or the wiring rerouted until the interference
is eliminated.
A more prevalent type of sync cross talk occurs between two non -
synchronous sources, such as local and network signals or local and video-
tape signals. This trouble is evident as a vertical bar or line moving non -
synchronously back and forth horizontally on the video monitors when the
network or television -tape signal is fed to the program line. (Of course,
this will not occur if the tape is operated in the Intersync or Pixlock mode,
because the tape output then is in phase with the local sync generator.)
The condition is caused by cross talk between the local sync and the non -
synchronous tape or network signal.
This trouble is usually the result of ground loops. A ground loop in an
otherwise well designed installation is most often the result of an open or
intermittent ground at one end of a coaxial cable. When an open occurs,
the signal at the open-shield end must obtain its ground return through a
number of racks. Each cable should be disconnected from the sending end
and checked with an ohmmeter from center conductor to shield to deter-
mine if the termination resistance is obtained. If the shield is open, no
continuity will exist. Always twist both the sending and receiving con-
nectors while making this check so that loose, intermittent, or high- resist-
ance connections will be made evident.
More recent sync generators, particularly those of modern digital design,
are normally quite stable and trouble -free compared to the older types. The
stability of the color -subcarrier generator is many times better than that of
older systems, and this is particularly important for second- and third -
generation dubbing on color tape systems.
12 -4. CHECKING VIDEO CHARACTERISTICS

Knowing the normal gains of individual amplifiers (such as preampli-
fiers, modulator video stages up to the frequency -modulated stages, rf
stages from modulated stage to modulator output, demodulator unit, etc.)
is an important factor in preventive- maintenance procedures. This step in
preventive maintenance will help in keeping the system signal -to-noise
ratio within specifications and, in tube -type systems, will indicate when the
need arises for carrying out tube -transconductance checks and making
replacements.
As a specific example, measure the video-head output (with a scope
directly at each commutator brush) on playback of a standard tape such as
the alignment tape. This rf output is normally from 2 to 5 millivolts. Then
measure the output of the individual head preamplifiers, and compare this
peak-to -peak value to the normal value, determined when the system is
known to be properly functioning. All of these values should be recorded

and available to the maintenance personnel.
The general condition of the video -processing amplifier can be inter-
preted if the normal range of input level is known. An example of a
specific tabulation follows:
Normal input: 1 volt (p-p)
Minimum input: 0.5 volt (p-p) for same signal -to -noise ratio
0.25 volt (p -p) just above sync breakout
(A) Proper proportionment at input. (B) Acceptable reproduction of A.

Fig. 12 -3. Multiburst signal as proportioned for older tape systems.
Use of Test Signals

The need for attention is most quickly checked with the keyed -burst
signal, as illustrated in Fig. 12 -3. On all older tape systems, care should be
taken to proportion the amplitudes of the single-frequency bursts as shown
for the input signal of Fig. 12 -3A. The energy of frequency components
in the normal monochrome picture is quite low above 1.5 MHz. If sine
waves are fed into the system with the higher frequencies at the same
amplitude as the white -pulse reference, considerable high - frequency power
exists, and high- frequency overload of the modulator due to the previously
discussed pre-emphasis may occur. The same type of overloading occurs
in the demodulation process due to natural limitations in high- power, high -
frequency linearity. The response obtained from such adjustment of the
input signal will indicate some loss of higher frequencies as compared
with Fig. 12 -3B. Also, a shift in the ac axis occurs, indicating selective -
frequency clipping. This type of input signal exceeds the normal limits of
the modulation -demodulation process and is of little value, except as a
safety- margin check. The multiburst signal generator should be carefully
designed so that it does not have a second- harmonic content. The pattern
will give the appearance of poor frequency response on playback at any
frequency where the second- harmonic content is high. Many professional
multiburst signal generators employ a 4.5 -MHz filter at the output, where
the highest burst frequency in the signal is 4.2 MHz.
The above situation does not hold true for modern low -band and high -
band systems, such as the Ampex VR -3000 and AVR -1 or RCA TR -70
systems. The normal multiburst test signal is used for these systems, as
illustrated in Fig. 12 -4. The extended system capability is brought about
(A) Full- amplitude multiburst signal (B) Playback of recorded signal on

input for recording. low -band color standards.
Fig. 12 -4. Normal multiburst test signal.
by the use of double sidebands in the fm -video system and modulation -

demodulation process.
NOTE: It is important to remember, however, that in the interest of
minimum phase distortions and maximum signal -to -noise ratio, the actual
video output response may be down as much as 3 dB at 4.5 MHz relative
to 1 MHz. This is due to the critically shaped response of the filtering
networks in modern tape systems.
Fig. 12 -4B shows the output of a modern low -band color tape system.
The reproduction of the full multiburst test signal by a modern high -band
system is practically indistinguishable from the input signal in a photograph.
If a video sweep is used for fm circuit alignment, remember that the
highest video frequency (4 MHz) is represented by a carrier frequency of
1 MHz (lowest carrier frequency) Thus, peaking the carrier at the lower
.
end increases the response at the higher video frequencies.

When it is necessary or desirable to insert a test signal into the video pre-
amplifiers, use a wideband video pad such as the one shown in Fig. 12 -5.
Coaxial Connectors
Test Signal -. -1.- To Preamp Input
Fig. 12 -5. A 46 -dB video pad.

The pad is inserted between the test signal generator and the preamplifier
input by means of short coaxial cables. The test signal generator should be
adjusted for a signal output of no more than 0.5 volt (p -p) , to insure that
the 46 -dB attenuation will prevent any possible overloading of the pre-
amplifier.
Inability to bring a tube -type video amplifier into proper amplitude -
frequency response specifications by adjustment of the appropriate peaking
circuits may be due to any one or a combination of the following faults:
1.Changed plate load. An increase in the plate load will reduce the
response at higher frequencies. A decrease in the plate load will
reduce the gain and increase the response at higher frequencies.
2. Defective peaking coil or swamping resistor across the coil.
3. Low- transconductance tubes. (However, in certain types of negative-
feedback amplifiers, loss of highs may result from tubes that show
normal transconductance on a tube checker. Always replace tubes in
negative- feedback amplifiers before making further checks for troubles
when video-response checks indicate loss of high- frequency response.)
Always bear in mind the turn- around in fm peaking circuitry due to the
nature of the modulation system. In this case, for example, an increase in
plate load, causing reduced response at the higher carrier frequencies, re-
sults in a loss of low- frequency video and an increase in high- frequency
video.
The evaluation of low- frequency response is important in determining
absolute values of picture streaking. It is important to understand that the
degree of streaking observed on a picture monitor depends not only on the
monitor -amplifier characteristics, but also on the ratio of the brightness-
and contrast -control settings. There is almost always some visible streaking
when a window signal, or any other white bar on a black background, is
displayed if the bar extends over an appreciable portion of the scanning
line. This was the primary reason for development of the white -window
test signal, so that a truly accurate measurement can be obtained from the
cro presentation in quantitative terms.
The phase characteristic related to the low -to-high frequency response
ratio is closely related to the absolute measurement of the low- frequency
response. The shape of the passband response curve determines the tran-
sient response of the system. Depending on the rise time of the window
signal, an indication of transient response may be determined to some
degree, but this characteristic is more accurately shown by the sine pulse.
NOTE: The makeup and practical measurements of the sine- window
signal and the 20T modulated test signal were covered in Chapters 2 and
8 of Television Broadcasting: Systems Maintenance,1 and will not be re-
1Harold E. Ennes, Television Broadcasting: Systems Maintenance (Indian-

apolis: Howard W. Sams & Co., Inc., 1972).
100 E
80
60
40
I Back Porch Level
-20
/
I
J L
I I
\ -l- 1
1
m
20 I I I 8c I I I
-lcmor0.1251
Limits -Inner: - 2% Outer: K ^ 4%
K
Window and 0.125 us T Pulse Response
...L- -40
(A) 0.125H /cm sweep rate.
NJNVIlN!!.
80
60
7cm or 0.175H
3cm
40
1 cm or 0.025H 3cm or 0.075H 2cm or 0.05011-
20
J L
7.5-
20
Limits -Inner: K - 2% Outer:

Window and 0.125ps T Pulse
K - 4%
R sponse
- f
40
(B) 0.025H/cm sweep rate.

Fig. 12 -6. Use of sweep magnification
peated here. The student or practicing engineer not familiar with such
techniques should review this material at this time.
In any case, the Tektronix Type 529 waveform monitor is commonly

used in all vtr installations, and we will briefly cover the use of this
monitor before proceeding.
When the DISPLAY switch is set to 0.125H /cm, horizontal-sync timing is
measured in terms of H ( the time between horizontal -sync pulses, or the
time from the start of one horizontal line to the start of the next line) .
When the (HORIZ) MAG switch is set to X 1, one complete horizontal line
is displayed in a sweep length of 8 cm (see Fig. 12 -6A) In addition, .
refer to Table 12 -1, which lists the sweep -time relationship between H /cm
and microseconds /cm.
NOTE: The 2 Field position is the only position of the DISPLAY switch
for which a free -running trace can be obtained in the absence of a trigger
signal. Thus, the 2 Field position is useful for determining whether the
video signal is absent.
For pulse measurements at 0.125H /cm, one centimeter equals 0.125H.

For example, in the NTSC signal specifications, 0.125H is the minimum
time interval between the leading edge of the horizontal -sync pulse and the
end of the color burst.
K - Factor of Pulse-to-Bar Ratio

K- Factor of Bar
100
80
PO
K- Factor of Pulse
40
20
20
Limits -Inner: K - 2% Outer: K - 4%
Window and 0.125µs T Pulse Response
-40
(C) 0.005H/cm sweep rate.

in display of horizontal waveforms.
Table 12 -1. Sweep -Rate Settings
DISPLAY Switch Setting

(Horiz)
Mag 0.125 H/cml 0.250 H/cml
X 1 0.125 H/cm 0.250 H/cm

7.94 µs/cm 15.9 µs/cm
X 5 0.025 H/cm 0.05 H/cm
1.59 µs/cm 3.18 µs/cm
X 25 0.005 H/cm 0.01 H/cm
0.318 µs/cm 0.635 µs/cm

Also applies to the Line Selector settings of the DISPLAY switch.
If the (HORIZ) MAG switch is set to X5 when the DISPLAY switch is at

0.125H /cm, the time -base sweep rate is 0.025H /cm. This sweep rate is
used, for example, to make horizontal -sync pulse- waveform measurements
such as those shown in Fig. 12 -6B.
If the (HoRIZ) MAG switch is set to X25 when the DISPLAY switch is
at the 0.125H /cm position, the sweep rate is 0.005H /cm. This sweep rate
is useful for measuring the rise time and fall time of the horizontal -sync
pulses, to count the cycles of color burst (Fig. 12 -6C) , and to examine
portions of a complete line.
The 0.25H /cm position of the DISPLAY switch is another calibrated
sweep rate which is useful for making horizontal line and sync -pulse wave-
form measurements. In the X 1 position of the (HORIZ) MAG switch,
approximately 21/2 horizontal lines are displayed. Table 12 -1 lists the
sweep -time relationship between H /cm and microseconds /cm for each
magnifier switch position.
When the DISPLAY switch is set to either of the Line Selector positions
and the LINE SELECTOR control is used, it is possible to range into the top
of the picture to examine any one or two lines, depending on whether the
DISPLAY switch is set to the 0.125H /cm line selector or 0.25H /cm line
selector position. Also, the LINE SELECTOR control can be set so the portion
of the vertical -blanking pulse which may contain vertical- interval test
signals can be examined in detail. The 0.25H /cm line selector position,
in particular, is useful for observing sine pulses.
The range of the LINE SELECTOR control is such that any portion of field
1 or field 2 can be examined. Either field 1 or field 2 is selected by means of
the FIELD switch. A special bright-up circuit in the Type 529 increases the
crt writing rate in either of the two line selector positions.
The graticule lines noted in Fig. 12 -6C are used to measure the sin2-
window signal. The K- factor evaluation is a simultaneous measurement
of three characteristics, as follows:
1. K- factor of pulse measures high- frequency characteristic.

2. K- factor of bar ( window) measures low- frequency characteristic.
3. K- factor of pulse -to -bar ratio measures midfrequency characteristic.
The outer limits in each case are marked with solid lines and designate
a K- factor of 4 percent. The inner limits for each case are marked with
dash lines which indicate a K-factor of 2 percent.
Although the graticule specifies a 2- percent and 4- percent K- factor for
a 0.125 -µs T- pulse, the same graticule is used for the 2T pulse by selecting
the appropriate time base on the Type 529 monitor. A time base of
0.125H /cm X25 is used for the T pulse, whereas the 0.25H /cm x25 time
base is used for the 2T pulse. The 2T pulse (0.250 -µs half -amplitude dura-
tion) is normally used for video-tape systems measurements. Fig. 12 -7 illus-
trates the reproduction of the 2T-window test signal of the high -band test
tape mentioned in Chapter 11. For proper measurement, the sine pulse
should be adjusted by scope gain to reach the top of the bar signal (100 -
percent reference amplitude) as shown. The pulse should then fit well
within the 2- percent K- factor lines (dash lines) on a properly functioning
high -band tape system.
Top of Bar IWindowl Signal 20T Modulated Pulse WindowlBarl 2T Pulse
.o.
aa-...nM ...
Fig. 12 -7. Waveform- monitor display

.
.,ft... . m
Fig. 12 -8. Playback of 2T- pulse, 20T-

during playback of 2T- pulse, pulse, and window signal from
high -band standard. high -band test tape.
Fig. 12 -8 shows the reproduction (playback) of the 2T pulse, 20T

modulated pulse, and window signal of the high -band test tape described
in Chapter 11. A somewhat expanded time base was used here to make the
reproduction of the 20T modulated pulse more apparent. The T and 2T
pulse graticule does not measure the 20T modulated pulse by K- factor lines.
In general, the reproduction of a properly operating high -band tape system
should have response to a 20T modulated pulse within 2 percent of the
undistorted waveform of Fig. 12 -9A. (In Fig. 12 -9, T indicates the top of
the window signal.)
Pure linear amplitude distortion is shown by Figs. 12 -9B and 12 -9C, and
pure delay distortion is shown in Figs. 12 -9D (3.58 -MHz component
delayed) and 12 -9E (3.58 -MHz component leading). Note that in all four
(A) Undistorted modulated 20T- pulse.
(B) Decreasing amplitude with

increasing frequency.
d111,
dl
(C) Increasing amplitude with

d2
dl (D) Increase in group delay with

d2
(E) Decrease in group delay with

d2
T
(F) Simultaneous amplitude decrease and

/1 increase in group delay at 3.58 MHz.
(G) Simultaneous amplitude increase and

decrease in group delay at 3.58 MHz.
Fig. 12 -9. Typical distortions of

modulated 20T- pulse.
waveform distortions, dl = d2. However, when both amplitude and delay

inequalities occur simultaneously, the resulting baseline distortion is not a
linear addition of the two components, as shown by Figs. 12 -9F and 12 -9G.
Techniques for rf and video sweeping are thoroughly outlined elsewhere.2
In tracking down amplitude -vs- frequency troubles or transient -response
problems in the modern tape system, the techniques are the same, but the
equipment required is different. As described previously, the fm portion of
a high -band recording system requires a sweep width to at least 30 MHz as
compared to the more conventional 10 MHz used for normal video sweep-
ing procedures. In addition, all modern tape- system fm and filtering cir-
cuitry has factory- adjusted components that are sealed after alignment.
Unless proper equipment is available, the field service department of the
tape-system manufacturer should be contacted for rf -video alignment ad-
justments after necessary replacements of critical components. Playback of
the test tape, followed by recording and playback of appropriate test signals
as previously outlined, will reveal whether such alignment is necessary.
For differential -gain and differential -phase measurements, either the
stairstep with a superimposed 3.58 -MHz signal is used (as in Fig. 12 -6A) ,
or the sawtooth with 3.58 -MHz signal (Fig. 12 -10A) is employed. Fig.
2Harold E. Ennes, Television Broadcasting: Systems Maintenance (Indian-

apolis: Howard W. Sams & Co., Inc., 1972).
(A) Without high -pass filter.
(B) Through high-pass filter.
Fig. 12 -10. Playback of sawtooth portion of high -band test tape.

12 -10A shows the crt presentation for playback of the high -band test tape
mentioned in Chapter 11. Fig. 12 -10B shows the same presentation with
the Type 529 monitor placed in the high -pass position, with additional
gain used to bring the maximum portion of the envelope to 100 IEEE
units. Thus if a = 100 and b = 95, there is 5 percent differential gain at
3.58 MHz, white compression.
IMPORTANT NOTE: A rather complete study of differential gain and
differential phase at 3.58 MHz, as well as techniques for using the vector -
scope for differential -phase measurements, was presented in Chapter 8 of
Television Broadcasting: Systems Maintenance.3 Techniques and inter-
pretations are identical for the television tape system. Tests should be run
at 10 -, 50 -, and 90-percent APL for both differential gain and differential
phase, as fully covered in the above -mentioned reference. In video tape
systems, differential gain and phase can be caused by the head or head -
resonance adjustments. In the Ampex AVR -1 system, differential gain
controls are included in the playback amplifiers.
An example of a handy form to be used for differential -gain and differ-
ential -phase measurements is presented in Fig. 12 -11. In this example,
APL
Record Playback
APL APL
On On
10% 50% 90% 10% 50% 90%
A
A B
A
B B
A
C B
Fig. 12 -11. Example of form for differential gain and phase measurements.
three record /reproducers are included in the complete system. The test
signals are first recorded on machine A, and then played back on all three
machines. The performance is recorded on the form. This process is re-
peated for all three machines. This technique gives the maintenance tech-
nician a good indication of performance of each system, as well as verifica-
tion of having all machines adjusted to a "standard" of operation.
Signal -to -Noise Ratio
One video measurement now remains: the video signal -to -noise ratio.
Measurement of recording -playback signal -to -noise (S /N) ratio is a good
indicator of the need for a complete run -through on tube replacements
(on older tube -type systems, or in modern record -playback amplifiers using
3Ennes, loc. cit.

tubes), amplifier gains, input source and output lines, and demodulator
limiting circuitry. Some modern systems still employ vacuum tubes in the
record function as head drivers, and in the playback function as the first
preamplifier stages where nuvistors may be employed. All other circuitry
is normally solid- state.
Three generally satisfactory methods of measuring video-tape S/N ratio
will be described: (1) the oscilloscope method, (2) the vtvm method,
and (3) the Rohde & Schwarz UPSF video noise meter method. Method
1 is less time consuming and most convenient, but less accurate than
methods 2 and 3. Method 3 is the most accurate and is now being used
by the equipment manufacturers for factory tests.
The Oscilloscope Method-The tape used should not be completely new
and unpolished, nor old and worn or scratched; a tape should be used
that is in the condition employed for normal satisfactory recording and
playback in the daily schedule. The heads should be optimized, the de-
modulator properly balanced, and critical (optimum) tracking adjust-
ments made. To make the readings significant, the signal -to -noise ratio
should be determined on a peak -to-peak video to rms noise basis.
A. The test signal can be a "doorstep" ( three steps: one at black, one
at gray or midway between the black and white peaks, and one at
reference white level) or the standard stairstep signal described
before. Many prefer the "one- line -in -six" stairstep signal (Fig.
12 -12A) adjusted for a 50- percent duty cycle. This signal provides
a convenient method of swinging the modulation over the refer-
ence frequencies to provide the reference level, while simulta-
neously providing for a gray -step measuring point.
B. With the system optimized for recording, make about 5 minutes of
recording of the test signal.
C. With the system optimized for playback, connect an oscilloscope at
the demodulator output, and observe the playback of the preceding
test signal.
D. Fig. 12 -12B illustrates a typical playback of the test signal of Fig.
12 -12A. Note the additional thickness of the gray step. Adjust the
video output level to obtain the standard 0.714 volt (p -p) of video.
NOTE: Measurements of S/N ratio are made at any point ahead of the
ATC and processing amplifiers, to avoid false readings from regenerated
sync and spurious signals from any random ATC operation.
E. Expand the vertical deflection by using maximum scope gain. Read

the peak -to -peak excursion of the noise (at the 50-percent steps)
with the scope on wideband (10 -MHz) response, if the Tektronix
Type 524 scope is used. Assume for the moment that the measured
peak -to -peak value is 100 millivolts (0.1 volt)
.
F. To convert the 100 millivolts from peak -to -peak to rms value,
multiply by 0.35:
100 X 0.35 = 35 millivolts
= 0.035 volt
G. The voltage ratio is now 0.714/0.035, or 20.4. The dB equivalent
for a voltage ratio of 20.4 is 26. This is on a peak -to -peak video to
rms noise basis.
H. To convert the 10 -MHz bandwidth reading to a 4 -MHz bandwidth,
which is the useful passband of the video information, add 7 dB to
the above computation:
26 dB +7 dB = 33 dB
If a 4 -MHz scope amplifier or preamplifier is used, delete Step H. The
bandwidth is given in terms of the 3 -dB -down point. Note also that from
the analysis, 100 millivolts of noise is approaching the maximum allow-
able to meet specifications for a 35 -dB signal -to -noise ratio. Thus, the
100- millivolt (p -p) noise level becomes the warning flag on older low -
band systems. Newer systems achieve an S/N ratio of 40 to 46 dB. Also,
when later Tektronix scopes, such as the Type 547 (50 -MHz bandwidth) ,
are employed, the correction factor becomes 12 dB rather than 7 dB.

Due to the limitations in accuracy of reading the noise peaks on the
oscilloscope trace, the preceding method is subject to a variable error, de-
pending entirely on the operator's care and his familiarity with his par-
ticular scope-amplifier characteristics. The result under optimum condi-
tions should be within a few dB of the actual signal -to -noise ratio.
The VTVM Method, Older (Low -Band) Systems -The vtvm method
described here is more accurate than the scope method and is limited only
by the accuracy of the instrument used.
(A) The equipment required includes a bandpass filter (Fig. 12 -13

or equivalent) and a 4 -MHz vtvm, such as the Hewlett -Packard
Model 400D or 400H, the Balantine 314, or equivalent.
(B) Adjust the detector balance control on the demodulator for maxi-
mum rejection of carrier 10 -MHz ripple, with the modulator
carrier frequency set to 5 MHz and the deviation set at zero
(modulator input level control completely counterclockwise) .
Then raise the input level control for Step C.

(C) With a standard window or stairstep signal applied to the input
of the modulator, set the carrier frequency and deviation in ac-
cordance with the proposed SMPTE recommended practice, i.e.,
5 -MHz carrier frequency corresponds to blanking level and 6.8
MHz corresponds to peak white (low -band) .
(D) Record a two-minute section of window or stairstep signal, and
rewind the tape to the beginning of the recorded section.
. .
. .
. .
.
MINIM /NM
.
if u te o'
I 1 .
(A) Input to tape recorder. (B) Demodulator output.
(C) Expanded portion of 50% point.
Fig. 12 -12. One -in -six stairstep signal, 50% duty cycle.
(E) Place the machine in the reproduce mode and adjust the video
output level control on the demodulator for an output level of
0.7 volt peak -to-peak video, not including sync.
(F) Remove the video input cable from the input to the modulator
in order to disable clamping. Set the carrier frequency control
for a carrier frequency of 6.0 MHz, corresponding to a medium
shade of gray.
(G) Record a two-minute section of undeviated 6.0 -MHz carrier and
rewind the tape to the beginning of this recording.
(H) Remove the video cable from the demodulator output, and con-
nect the special bandpass filter to the output of the demodulator.
Connect the output of the filter to an rms- measuring vtvm.
(I) Place the machine in the reproduce mode; set the scope -selector
switch on the left-hand control panel to the switcher output posi-
tion, and adjust the tracking control for maximum amplitude of
the scope presentation.
(J) Read the value of rms noise on the vtvm, and compare this fig-
ure to the 0.7 -volt (p-p) level. Remember to take into account
the measured insertion loss of the filter (approximately 5 dB for
the filter shown in Fig. 12 -13) . Twenty times the log of this
ratio is defined as the video signal-to -noise ratio of the television
recorder.
NOTE: Later systems (high -band and low -band) employ built -in facilities
as described following the discussion of the noise-meter method.
Video Noise Meter Type UPSF -A

block diagram of the Rohde &
Schwarz type UPSF video noise meter is shown in Fig. 12 -14. This instru-
ment serves to measure unweighted and weighted noise voltages in tv

transmission systems. It is particularly suitable for measurements in the
region between the black and white levels of the video signal range, even
in the presence of line or field blanking intervals with or without sync
pulses. For this reason, the noise meter features horizontal or horizontal
plus vertical -interval blanking. Thus the test signal is cleared of the sync
and blanking pulses and, during the blanking interval, brought to the
mean value of the amplitude in the video signal range (Fig. 12 -15). The
loss of noise energy during the blanking period is taken into account
when calibrating the indication.
This noise meter was designed to satisfy the relevant CCIR recom-
mendations, Geneva, 1963 (Rec. 421.3.3) , concerning noise -voltage mea-
surements in tv transmission systems. A low -pass filter built to CCIR
recommendations (Rec. 421 Annex II) prevents noise voltages exceed-
ing the upper frequency limit of the transmission system from being
measured. A continuous -random -noise weighting network built in con-
formity with CCIR specifications (Rec. 421 Annex III) simulates the
sensitivity characteristic of the eye to noise voltages.
The video noise meter is suitable for measuring noise voltages of equip-
ment such as tv cameras, film scanners, and video tape recorders, and for
noise- voltage measurements on radio links, coaxial lines, tv transmitters,
tv receivers, and tv translators. In the case of noise -voltage measurements
LI
1
L2 L3 L4
I '1
47 pF 39 pF 22 pF
0.01 pF
Input
(Output of Recorder) 570
=68 pF *-1( 20pF
J To VNM
130 pF1 12P pF
75 22 p 1K
(A) Circuit.
North ill Coils Crest IMicrotranI Coils Frequency Relative Attenuation (dB)
Coil Part No. Value in pH Part No. Value in pH <100Hz 45
LI 1000 -F 14.5 -22 200 -3 6.6 -20 <10kHz 6
L2 1000 -1 55 -110 200 -5 33 -110 10kHz-4MHz 0
L3 1000 -H 29 -55 200 -4 16 -55 4.2 MHz 6
L4 1000 -F 14.5 -22 200 -3 6.6 -20 4.5 MHz 40
(B) Coil values. (C) Characteristics.

Fig. 12 -13. Bandpass filter (based on data originally supplied by Ampex).
Fig. 12 -14. Block diagram of video noise meter Type UPSF.

on color tv systems, a trap tuned to the frequency of the chrominance

subcarrier prevents the measurement of any residual components of the
subcarrier. Apart from the rms -value indication recommended by the
CCIR, the video noise meter also gives peak -to -peak value indication
capable of responding to noise peaks even of short duration.
The noise signal, cleared of sync pulses and blanking intervals, is avail-
able at a socket for the connection of an oscilloscope, so that a quasi -
peak -to -peak amplitude display can be obtained.
1110111
Video Signal (Gray Level)
With Noise Signal rw
so
Cancelling Pulses
7I
I i
Noise Signal With

Blanking -Interval -I I
Compensation I i
i
i
Fig. 12 -15. Principle of noise -voltage measurement with horizontal or

vertical -interval blanking.
The measurement and frequency ranges and the possibility of switching

over to different inputs and outputs as well as to internal or external
synchronization make the instrument suitable for a number of applica-
tions. Thus the cost of using the instrument for video -tape systems can
be justified by its other uses and applications at the studio and transmitter.
As mentioned previously, the Ampex VR -2000B and RCA TR -70 sys-
tems employ built -in facilities for checking the S/N ratio. These specific
applications follow.
Ampex VR -2000B S/N Ratio Measurement -Each VR -2000B system
is factory adjusted to assure that it operates well within the performance
specifications. These specifications include an E-E signal -to-noise ratio of
60 dB or better and a playback signal -to -noise ratio that is specified for
each recording standard in Table 12 -2. Because measurement of system
signal -to -noise ratio is occasionally a matter of considerable interest, a
facility for this measurement has been provided. The procedure which
follows explains the use of this facility.
QI "ADRUPLEX TAPE SYSTEM MAINTENANCE 415
Table 12 -2. Ampex VR-2000B Video Signal /Noise Specifications

Standard Minimum Signal /Noise Ratio'
Low-Band Color 40 dB
High -Band Color 46 dB
'Ratio of p -p noncomposite video to rms noise; noise measured at midgray level using wide
band true -rms voltmeter through 5.8-MHz low -pass filter; standard tip engagement.
(A) Set the AFC /CLAMP switch at one of the gray -level positions (i.e.,
position 3 or 4). Turn the VIDEO INPUT LEVEL switch to the off
position. Set the SERVO MODE selector to the Preset or the Line
Lock position. Thread a blank tape on the tape transport.
(B) Initiate the record mode, and record the gray -level frequency for
2 or 3 minutes; then press the STOP push button.
(C) Turn the TRANS SUPP switch (on the setup control panel) to the
CW setting. Connect the video-signal system output through a
5.8 -MHz low -pass filter to an rms- indicating meter. The rms
meter will indicate the E -E signal -to -noise ratio, which should be
60 dB or better. Reproduce the recording made in step B. The
rms meter will now indicate the rms noise content of the repro-
duced signal. The meter indication may then be compared with
the 0.7 -volt (or 0.714 -volt, depending on the recording standard)
normal level of the video portion of the composite signal, to de-
termine the S/N ratio.
NOTE: The CW setting of the TRANS SUPP switch causes the transient
suppressor to produce transient -suppression pulses that overlap the
interval during which the rf switch is transferring the pickup to succes-
sive video heads. Necessarily, these pulses occur at the horizontal rate as a
function of the phase of the head drum. Thus, this facility eliminates
head -switching transients which would otherwise greatly increase the rms
noise indication and produce an erroneous measurement of signal- system
noise.
There are inevitably many potential generators of noise in a complex
electronic system. (Amplifiers and random thermal effects are examples
of noise generators.) In the VR-2000B system, generated noise is ex-
cluded from the video -signal system because its input is not connected to
the output of the video heads until they have engaged the tape. At all
other times, the video -signal system receives incoming video from the
output of the modulator. This exclusion of generated noise eliminates the
time that would otherwise be required for the video electronics to first
recover from a noise condition when the play mode is entered. Thus there
is no normal condition in the VR-2000B system under which noise is
passing through the video signal path.
(D) Press the STOP push button.

4i6 TELEVISION BROADCASTING
RCA TR -70 S/N Ratio Measurement -The video signal -to-noise ratio
is defined as the peak -to -peak video voltage divided by the rms noise volt-
age. (Video here does not include sync. The noise is random noise only
and is not meant to include other undesired components such as switching
transients, cross talk, moiré, tape scratches, or similar items.)
The machine is first set up to record and play normally, with particular
attention on fm deviation and output video level. The normal level of the
video portion of the signal is 0.714 volt peak -to -peak for a one-volt com-
posite output (100 IEEE units of video and 40 IEEE units of sync) This.
0.714 volt is used as the video signal level for S/N calculations. The ma-
chine output is adjusted to give exactly this level as measured on an
oscilloscope.
To read the noise level separate from the video, a test recording is made
in the noise test mode, thus giving a recording of unmodulated fm car-
rier. The machine is then played back using the noise test mode, and the
demodulated noise output of the machine is measured with a meter. The
ratio of 0.714 to the noise reading gives the numerical signal -to -noise
ratio. For example, a noise reading of 0.00714 volt would give an S/N
ratio of 100 times (40 dB) .
The test recording for the noise measurement is made with the fm car-
rier at a frequency corresponding to picture gray level. The noise measure-
ment is made at a point in the system just ahead of the ATC and process-
ing amplifier. This is done to avoid false readings due to regenerated sync
and spurious signals due to random ATC operation.
In order to obtain a representative measurement of the video S/N
ratio, the operating conditions must be carefully controlled. The tape must
be a high -performance type and in good condition. The meter and oscillo-
scope must be carefully calibrated. The meter must be a true rms- reading
meter such as the Hewlett- Packard 3400A. Capstan tracking must be
carefully adjusted, and the guide position must be carefully set to mini-
mize switching transients. Another factor to consider is the condition of
the heads. The S/N ratio improves as the heads wear and the pole -tip pro-
trusion decreases. Of course, when the heads wear out, the S/N ratio rap-
idly gets worse.
If a Hewlett- Packard 3400A meter is not available, an approximate
measurement can be made with a Hewlett- Packard 400D meter if 2 dB
is added to the noise reading (subtract 2 dB from the S/N ratio). If no
meter is available, an approximate reading can be made by using the oscil-
loscope to measure the noise. In this case, peak -to -peak noise should be
measured. It has been found experimentally that, under these conditions, a
suitable conversion factor is 12 dB. Thus, divide 0.714 by the peak -to-
peak noise as seen on the oscilloscope. Express this ratio in decibels and
add 12. The noise reading should not include extreme noise peaks.
Proceed as follows to perform signal -to-noise measurements on the
RCA TR -70 system:
1. Set up the machine normally, using a multiburst or color -bar signal,

a good headwheel panel, and good, high -performance tape.
2. Be sure that the following procedures have been carefully followed:
(a) Deviation adjustment
(b) Record -current optimization
(c) Guide adjustment on test tape
(d) Playback equalization
(e) Demodulator balance ( performed on low -band monochrome
standard, then switch back to desired standard)
3. Switch the RECORD REF control on the reference generator module
to House Sync and the selector switch on the tape sync processor
module to TW.
4. Switch the guide servo module to manual.
5. Remove the MATC video module, and insert the signal module.
6. Connect the meter to pin 17 of the signal module using 75 -ohm
cable with a termination at the meter.
7. In the E -E mode, measure the blanking -to-white video level at the
meter termination using a carefully calibrated oscilloscope. This
level should be 0.714 volt peak -to -peak. Adjust the VID GAIN con-
trol on the postemphasis module for exactly 0.714 volt peak-to -peak
using the oscilloscope.
8. From this point on, do not change the adjustment of the VID GAIN
control.
9. Make a one -minute recording of the multiburst or color -bar signal.
After one minute, let the machine continue recording. Set the TEST
SELECT switch on the fm test panel to Noise, and press the RECORD
switch and then the TEST switch.
10. Continue with this noise -test (gray -level) recording for two min-
utes.
11. Rewind and play the multiburst or color-bar recording.
12. Adjust the guide position and tracking carefully.
13. Check the playback equalization, and adjust if needed.
14. When the noise-test recording (gray level) comes up, switch the
meter to the -40 -dB scale and read the noise.
15. Observe the demodulator on the monitor, and carefully adjust the
guide for minimum switching transients. Raise the monitor con-
trast to make the transients plainly visible.
16. Observe the fm level on the cro, and adjust the tracking for maxi-
mum output.
17. Make further fine tracking and guide adjustments to minimize the
noise.
18. Use the minimum noise reading, but do not count occasional dips
which may be due to the meter time constant.
NOTE: For convenience in reading signal -to-noise ratio in decibels, note

that on many Hewlett- Packard meters 0.0071 volt is nearly in line with
-41 dB. On these meters, the S/N ratio can be read directly by subtract-
ing one. Thus -41 = -40 dB.
19. If the specified S/N ratio (same as in Table 12 -2) is not obtained,
perform the following steps to effect an improvement:
(a) Use another tape.
(b) Degauss the heads.
(c) Reoptimize the head currents.
(d) Observe the noise with an oscilloscope, and check for the
presence of hum, interference, oscillations, etc. The noise
should be random.
(e) Look at the E -E noise with the meter. To do this, use the same
procedure as for record - playback noise, but with the machine in
the stop and E -E modes, set the TEST SELECT switch on Noise,
and press the TEST switch. The E -E S/N ratio should be from
-55 to -60 dB.
(f) Use another headwheel panel.
(g) Change the nuvistors in the head preamplifier (playback) and
compare the S/N -ratio measurements.
Recording and Playback of Nonsynchronous Signals
Certain troubleshooting procedures require the use of nonsynchronous
signals which are recorded and played back as necessary. These non-
synchronous signals may consist of sweep, cw sine waves, or square -wave
signals not containing sync or blanking information.
Recording Procedure -To record nonsynchronous signals, proceed as
follows:
1. Check the level indication on the cro, and set the input level at its
source to provide an indication of 0.7 volt (100 IEEE units). This
is the correct input level for recording.
2. Set the TEST SELECT switch on the fm test panel to the Input posi-
tion.
3. Press the MASTER RECORD switch on the record panel; then press the
TEST switch on the fm test panel.
Playback Procedure -To play back nonsynchronous signals, proceed as

follows:
1. Set the TEST SELECT switch on the fm test panel to the Input position.
2. Press the PLAY switch on the play control panel; then press the TEST
switch on the fm test panel. The machine will now play back a
nonsynchronous recording.
NOTE: The TEST switch on the fm test panel must be pressed after going
into an operating mode, or the switching logic will inhibit the test func-
tion.
12 -5. USE OFAUTOMATIC COLOR

CIRCUITRY IN MAINTENANCE
The experienced maintenance technician soon learns to employ infor-
mation supplied by the Ampex Colortec and Velocity Error Corrector, or
the RCA Cavec, in judging certain system functions. Velocity-error and
chroma -error signals can be monitored for various defects in setup and
alignment.
For example, see Fig. 12 -16. On the RCA TR -70, when the cro CAC
monitoring push button is depressed, the upper and lower limits of the
automatic correction range are displayed as shown. Fig. 12 -16A shows
the pattern when the channel equalization controls are not matched (see
Fig. 11 -14A). Without automatic chroma correction, severe color band-
ing would result. With CAC in the circuit, the amount of banding shown
by Fig. 12 -16A would still be within the automatic correction range, and
no color banding would be apparent. Fig. 12 -16B shows the display when
the four channel playback amplifiers are properly equalized.
NOTE: The display described above is also available on the Ampex
AVR -1 when the CHROMA push button is pressed for the A-scope. Upper
and lower limits are not displayed on the Ampex, but if the burst levels
are correct, the burst -envelope display will be at the center line of the A-
scope.
Upper Limit
Excessive Banding but Within

Automatic Correction Range
Lower Limit
(A) Equalization controls not matched.
Upper Limit
}Optimum Equalization
Adjustments, All Channels
Lower Limit
(8) Playback amplifiers properly equalized.

Fig. 12 -16. CAC monitoring waveforms, TR -70 system.
Head -resonance compensation has been discussed previously. There is

a further adjustment that is particularly important to color, especially if
the Auto Chroma accessory is present in the Ampex VR -2000B system.
This is an adjustment of the master equalizer and the individual channel
playback equalizers to minimize color differential gain in all the channels,
and to verify the settings of record levels and playback resonance com-
pensators.
The front panel of the master equalizer (module 17 of the Ampex VR-
2000) includes two controls. One is potentiometer R1, MASTER PB RE-
SPONSE; the other is 5- position switch Si, TURNOVER FREQUENCY.
The procedure for the minimization of color differential gain makes use
of a test signal that is standard in many color- equipped stations. It is a
composite of a stairstep (or sawtooth) signal with subcarrier added. The
stairstep (usually 10 steps) is the luminance portion of the signal, and it
varies between the blanking and peak-white levels; the subcarrier fre-
quency is nominally equal to the color -subcarrier frequency of the record-
ing standard in use. For this test, the subcarrier should be adjusted to a
peak-to -peak amplitude equal to the peak -to-peak burst level and to the
peak -to -peak sync level (i.e., 40 IEEE units in the oscilloscope display) .
Color- equipped Ampex VR -2000B systems include an etched board
(mounted in the stability marker box) on which a subcarrier filter and
subcarrier detector are built. The output of the detector is routed to the
channel -B input of the waveform monitor. Thus, when channel B of the
waveform monitor is in use, the display is the rectified subcarrier of the
signal selected by the video monitor selectors (on the audio and video
monitor selectors panel) . If the Tektronix Model RM 529 waveform
monitor is employed, the SYNC switch must be set at Ext in order to obtain
a stable presentation. The procedure follows.
(A) To begin the procedure, record the standard test signal for sev-
eral minutes, taking care to set the luminance deviation correctly.
(B) While reproducing the recording just made, examine the repro-
duced waveform. The output of the demodulator may be exam-
ined in the display of the waveform monitor by pressing the
DEMOD output push button on the monitor switching panel. If
all the compensators are correctly adjusted, and differential gain
has been minimized, the display will be the normal rectified en-
velope of Fig. 12 -17. (The normal noise shown is inherent in
the record /reproduce process.) If the compensators are not set
correctly, differential gain will be present, and the rectified en-
velope will slope in a direction and at an angle that is directly
related to the accumulated adjustment error. The differential gain
of the off -tape signal should be compared with that of the input
signal by alternately pressing the DEMOD and INPUT push buttons
(of the video monitor selectors).
Peak
White
Level
Input Test Signal
Blanking Level
Output of Subcarrier Filter

(Available Only Within the
Detector Etched Board Assembly)
1-111111111111111111.111
Normal Noise
Output of Subcarrier Detector

(Normal Rectified Envelope)
Output of Subcarrier Detector

(Differential Gain Present)
Slope may appear in either direction, at any angle
Fig. 12 -17. Input test signal vs output of subcarrier filter and detector
Ampex VR-2000B system.
The controls that may require slight retuning to minimize differential

gain include MASTER PB RESPONSE (on the master equalizer) and play-
back equalization controls oil, cx2, cx3, and Cx4 (on the operating
control panel) . The TURNOVER FREQUENCY switch (on the master
equalizer) must be set at the optimum tap. Repeated tests on a series of
VR -2000B systems have shown that for operation on 525 -line high -band
standards, typical settings are:
TURNOVER FREQUENCY: Position 3 or 4

MASTER PB RESPONSE: "9 O'clock"
CH1, CH2, CH3, and 014: Between and -3 -4 on their scales.
Examination of the demodulator output may reveal that the outputs of
the video heads exhibit significant differences between channels. For ex-
ample, the output of one or more heads may exhibit a concave or convex
differential gain curve, or a slope instead of a flat response. This is evi-

dence that the FREQ COMP and Q COMP controls in these head channels
require very slight retuning. The necessary retuning should not exceed a
change of 0.2 on the FREQ COMP scale, or one numbered division on the
Q COMP scale. If retuning within these narrow limits does not provide
flat response in each channel, it is possible that the earlier adjustments
using the sweep generator were not made correctly and must be repeated.
It is important to remember that differential gain must be at minimum
at the time that the chroma level is correct. The operator will find that as
the individual channel equalizers are turned clockwise or counterclockwise,
the response line moves up or down and in so doing tends to tilt one way
or the other. The correct adjustment will have been achieved when the
rectified off -tape subcarrier waveform matches that of the E -E condition
in both height and flatness. A good way to test for this is to momentarily
depress and release the E -E push button (located under the system con-
trol panel) , which will temporarily return the system to the E -E condi-
tion, permitting the comparison. To prove that the adjustments are cor-
rect, a multiburst signal should be recorded and then reproduced. If the
reproduced signal exhibits flat frequency response and an absence of dif-
ferential gain, the adjustments are correct.
If a difference in differential gain between head channels persists, the
uniformity of the recording of individual channels should be checked as
follows:
(A) Turn all but one of the channel equalizers fully clockwise; leave
one at its normal setting.
(B) Set the TRACK SELECTOR at its Home position, and reproduce the
differential -gain test signal recording. Note the degree of differ-
ential gain of the channel selected in step A.
(C) Repeat step B with the TRACK SELECTOR set successively at 2, 3,
and 4. Determine whether or not a change of height or flatness
occurs. If each of the heads has been properly optimized in re-
cording, the differences observed will be negligible; if one of the
recorded tracks appears to differ from the others, that recording
channel should be reoptimized by the standard method previously
described. Table 12 -3 may be useful in determining which head
made the recording of a specific track.
12 -6. CHECKING SERVO STABILITY

The mark of a good servo system is immediate stabilization of the pic-
ture as the vacuum guide engages the tape with the rotating heads. Mo-
mentary instability immediately following this action is often caused by
the velocity loop in the head servo because this loop must function rapidly
to obtain control tight enough for the phase loop to take over. Aside from
Table 12 -3. Identification of the Video Head That

Recorded a Specific Track
Track Selector Setting

Playback Channel
Observed Home 2 3 4
- 1
2 2
1 4
3
2
1
3
3 3 1 4 2
4 4 2 3 1
"touchy" tubes (which usually can be located by light tapping with a

pencil) , the following points all have a bearing on general servo stability:
1. Check all tape -transport tension adjustments, cleanliness of the head-

wheel and capstan, and the control -track head.
2. Check the condition of the tape and the number of splices; the lat-
ter is particularly important if the video heads have been worn to
1 mil or less of tip projection.
3. Check the centering of adjustments; always be certain that frequency

controls are placed in the center of the range at the limits of which
frequency is lost. It is also good practice from a preventive- mainte-
nance standpoint to ascertain the minimum and maximum pulse
width obtainable from multivibrators. In this way, the time when
proper pulse widths or a square wave cannot be obtained may be
anticipated before trouble occurs (particularly in older tube -type
equipment) .
NOTE: In setting the free -running frequency of 240- or 60 -Hz oscil-

lators of the triggered type, use the power -line triggers on the scope, and
adjust the oscillator for slight left -to -right drift on the scope trace. This
sets the oscillator free -running frequency slightly lower than the trigger
frequency (accounting for the slightly lower than 60 -Hz field rate of
color) so that trigger control is stable. If the free- running oscillator fre-
quency is higher than the trigger frequency, unstable operation may
result.
4. Become thoroughly familiar with the normal flow path of all con-
trol pulses so that the inputs and outputs of the various individual
chassis or circuits can be quickly traced.
5. Check power supplies and power- supply regulation on a regular
basis.
6. A closed -loop servo system has two major functions, loop gain (can-
not exceed unity) and phase gain (cannot go to 180° ) If operation
.
does not remain within these limits, oscillation occurs. The load on
the headwheel or drum can be set on the heavy side by increasing

tip penetration and then on the light side by decreasing tip penetra-
tion to the point of loss of tape contact, ignoring jogs in the picture.
If a tip penetration is found where the worst possible stability oc-
curs, set the servo loop gains and phase gains for optimum stability
at this point. Return the tip penetration to standard. Stable opera-
tion should result throughout the life of this headwheel assembly,
barring other troubles.
7. Other units may affect servo stability. Check demodulator (detector)
balance, processing -amplifier stability (afc circuits) , and stability
of reference pulses relative to a source of known stability such as
the local synchronizing generator.
8. Difficulty in servo lock-up upon initial start of the machine (in play-
back) can be caused by an automatic chroma corrector that greatly
increases the chroma gain before lock-up is achieved. Logic circuitry
is normally a part of such systems so that automatic correction locks
out until stable playback is achieved. Always remove any automatic
correction circuitry where possible to check for any effect on servo
lock -up time.
12 -7. UNDERSTANDING SERVO FUNCTIONS

It is important to have a thorough understanding of basic servo func-
tions. Let us review these requirements briefly as follows.
The low -band monochrome, nonsynchronous (head-tachometer pulse)
mode of operation is a relatively "loose" head-motor servo. It is entirely
satisfactory for monochrome operation when it is not necessary to mix
other signals with the tape playback. But for color operation, even when
it is not required to mix a local signal with the tape- playback output,
time -base requirements include some form of "synchronous" operation.
Essentially what we must do is "tighten up" the servos, as described in
the following steps:
1. The field -rate playback signal must be made a standard color field
rate of 59.94 Hz. Therefore, there must be a comparison between
the field rate coming from the tape system and a standard field rate
supplied by the local sync generator locked to a color standard. Any
error between these two signals must be utilized to control the head
(and capstan) servo.
2. The horizontal -rate playback signal must be made a standard color
line rate of 15,734 Hz. Therefore, there must be a comparison be-
tween the horizontal rate coming from the tape system and a stan-
dard horizontal rate from the local sync generator locked to a color
standard. Any error between these two signals must be utilized to
control the head (and, in turn, the capstan) servo.
3. Now that the servos are being held to a basic color standard, the
"dot interlace" of the color -subcarrier information must be main-
tained in spite of the tolerances of the mechanical head-to -tape
contact. To do this, there must be a comparison between the color
sync burst coming from the tape and the color subcarrier from the
local color standard. This comparison occurs in the Color ATC sec-
tion of the video control circuitry. This comparison enables cir-
cuitry which provides a "regenerated burst" which can then be
phased for proper flesh -tone reproduction on the color monitor. If
you keep the time -base concept, you realize this is the "tightest"
servo and signal requirement for the tape system.
To Variable Delay
To Vert -Coincidence (Dependent on Coincidence
Detector and TVA Between Local and Tape To Head -Motor
Phase Comparator Vert -Rate Pulses) Control Circuit Adder
i 1
Field -
Low Freq High -Freq
Pulse
Er or Error
Gen
Horiz -Sync i___... Phase Horiz-Sync

Sep Comp Sep
1
Demodulated Local
Tape Sync Sync Gen
(A) Added circuitry for full genlocked control.
Trapezoid Formed From Head

Tachometer Pulse (Head Speeds
-- -Goes Positive
-Ref Volts
To
Voltage-Controlled
Head -Motor Drive
Goes Negative
Sampling Pulse
Sampling Pulse From:
RECORD MODE: Reference Vert From Stripped Sync of Video Signal Being Recorded
PLAYBACK MODE:
111 NONSYNCHRONOUS: Reference Vert From Local Sync Gen or Power Line
121 VERT ALIGN: Reference Vert From Comparison of Local Vert Rate to Demodulated-Signal Vert -Sync Rate
Note: This mode also involves control of capstan motor to "track' head playing vert sync to local vert sync.
131 GENLOCKED: Same as 121 but Added 'Time- Modulated" Vert Comparison to Horiz -Sync Errors
Note: This mode also involves additional error correction to head -motor control circuitry at horiz rate.
(B) Pulses in record and playback modes.

Fig. 12 -18. Headwheel control.
Now go to Fig. 12 -18A, which shows the basic added circuitry needed
to obtain full genlocked operation of the tape system. The demodulated
tape sync from the playback signal is phase- compared to local horizontal
sync. The error signal is split into two components: a low- frequency (or
phase) error which feeds the variable delay and a high- frequency (ve-
locity) error which adds to the headwheel -motor control circuitry.
The system incorporates a number of lock -sense buses. For example,
such a bus comes from the vertical-coincidence detector. When vertical
lock is achieved, the fully synchronous mode can take over as shown by
Fig. 12 -18B. Such circuits normally energize relays which sequence the
lock-up procedure.
Note that the primary signal for Fig. 12 -18B is the head -tachometer
pulse formed into a trapezoid. The sampling pulse changes through the
required modes of operation as shown. Step 3 is the final step, and the
tightest servo.
The simplest operation of the capstan servo occurs in the record mode.
In this mode, the capstan motor is electronically synchronized to the fre-
quency and phase of the head motor. In the playback mode, more recent
tape systems have added circuitry for faster lock -up and tight control in
synchronous operation.
Until the head motor is first locked up, the capstan must be driven
aimlessly from either a free -running multivibrator or an integration of
Edit Pulse See Text Tape Frame

Pulse
\
/Control
Head
-Track CT 240 -Hz
Pulse Gen
30 Hz
i
CT
Tape
Phase
30Hz Coincidence Sense

Tracking
Control
Ref Frame Local Composite Sync
Pulse IHoriz & Vert Locked
to Color Std)
Error Volts
Capstan Osc
IVCO' 240Hz
Nominal
60Hz Nominal
To CapstanMotor a4
Drive Amps Counter
Fig. 12 -19. Tightening capstan playback servo.

the speeding -up headwheel motor. But if the head -speed (tachometer) sig-
nal is removed from a requirement of initial capstan lock -up, this condi-
tion will not exist.
The vacuum tape guide does not engage the head with the tape until
about 11/2 to 2 seconds after machine start is initiated. But the tape is
engaged with the control -track head ( which is mounted on the vacuum
guide) at all times. This signal is available immediately upon start of
the machine.
Also available, even before start -up is initiated, is the reference frame
pulse formed from local sync. This is always present at one side of a phase
comparator, as illustrated basically in Fig. 12 -19.
Consider the simplest mode of operation, the nonsynchronous playback
(neither vertical nor horizontal lock -up with local sync) . Now follow
through slowly and carefully. You know that the tape frame pulse, since
it is derived from the demodulated playback signal, cannot be immediately
available. But the recorded edit pulse (if any) is available immediately.
Again, if you are following carefully and using what you know about
timing relationships, you know that the edit pulse has the same relation-
ship (during initial recording) to the head vertical sync as the tape frame
pulse would have. So if the edit pulses are available and usable, the system
will lock up vertically as in Fig. 12 -20B. But this is not a requirement for
nonsynchronous operation.
Reference Frame Pulse(30Hz)

CT f8 130Hz1
(A) Nonsynchronous mode of playback.
Reference Frame Pulse 130Hz1 Edit Pulse First 4 Secs

or CTe8 130Hz)
Then: Tape Frame

Pulse 130Hz)
(B) Vertical -align playback and synchronous mode.
Head Tachometer : 8 130 Hz)
CT'8130Hz)
(C) Synchronous playback, vertical-align requirement dropped.

Fig. 12 -20. Capstan control for faster lock -up.
The point is that if no edit pulses are available or if they are degraded,
the capstan can still lock up to the condition shown in Fig. 12 -20A. In
practice, the circuitry contains AND gates (coincidence gates) , OR gates,
and lock -sense buses to insert or remove reset -pulse gates.
The second step of lock -up is shown in Fig. 12 -20B. The trapezoid is
still derived from the ever -present reference frame pulse. The sampling
pulse is that just described for nonsynchronous operation. (The system
will already lock up vertically if edit pulses are present.)
After the head motor has been engaged by the vacuum guide, the tape
frame pulse (demodulated from the video vertical interval) becomes
available, and the synchronous mode of operation (as far as the capstan
is concerned) exists. It is now the job of the capstan to keep the head
playing back vertical sync coincident with local vertical sync. It will speed
up or slow down according to the command of the error signals in order
to maintain this tracking.
When framing ( vertical alignment) is not required, the head servo
drops the requirement for tva (tape vertical alignment) , and the variable
delay section is controlled only by the circuitry of Fig. 12 -18A. Then the
head motor, after a disturbance, locks immediately to the next horizontal
line (although a noncorresponding sync pulse is used) , and recovery is
much faster than waiting for vertical framing.
At the same time, the capstan must now follow changes in head phasing
(which occur after disturbances) to maintain the tracking set by the
control -track phase- adjustment knob (tracking control). So now the cap-
stan reference becomes that of Fig. 12 -20C. The servos are being con-
trolled by station sync, but you cannot count on being able to use special
effects, etc., since proper phasing may not exist with other local sources.
This is the Ampex horizontal mode, or the RCA Linelock mode.
12 -8. CHECKING THE SAMPLE -AND -HOLD CIRCUITRY
It should be apparent at this point that the sampling servo is quite

common in vtr systems, and the maintenance technician should become
familiar with techniques for checking and servicing servos. Remember
that even the Automatic Timing Correction (ATC) circuitry becomes a
part of the overall servo system, and that this circuitry also involves the
sample- and -hold technique.
Assume the picture is unstable, possibly tearing out, even in the non -
synchronous mode of operation. You might notice, by using the monitor-
ing selector switch, that this is first occurring at the Amtec or the ATC
output, while the demodulator output is satisfactory. Or, the condition
might be such that the servos will not lock at all, and a stable picture is
not available even at the demodulator output. Further assume that the
system has selectable time -error correction in or out, and that you find
that the picture stabilizes with ATC in the out (off) position. There
are two possibilities: (1) the timing is incorrect, or (2) the trapezoid,
sawtooth, or sampling pulse is missing or low in amplitude.
Fig. 12 -21 is a representative circuit to illustrate the basic testing tech-
nique. This is applied specifically to ATC, but it will work with any of
the sampling error circuits in the tape system. Use a dual -trace scope with
one input at the Q1 collector (sample pulse) and the other input at the
trapezoid connection to the center tap of T1. (A test point is normally
provided at such connections.) Check first simply for the presence of both
pulses. If one or the other is missing, check back through that particular
path.
Voltage-
Video In
MATC
H Variale
Delay Line
Out
(
(
Proper Timing
(Dual-Trace Scope)
_FL_
From I-3 is
Trap Gen
- Ref -4. Diff Amp Phase

Split Q2
Sample
Pulse
Error
1 Det
THA Sig al Out
Back to E ror Det Amps
(See Fig. 6 -2.)
Q3
Fig. 12 -21. How to distinguish between time error and amplitude error.
Proper timing is evident if the sample pulse occurs near the center of
the trapezoidal slope. If proper timing is not obtained at this input to the
error detector, the fault will be found in the horizontal- alignment differ-
ential amplifier. (This measures the phasing between sampling pulse and
trapezoid to hold the variable delay line in the center of its range, using
a dc reference voltage against which to measure.) If timing is correct up
to the input of the error detector, you know the fault then lies in the
error detector itself or in following amplifiers. You will be able to apply
this type of troubleshooting technique to any pulse -sampling afc circuit,
of which there are many in tape systems.
Some of the latest systems do not have ATC on or ATC off available on
a switch. However, if it becomes evident that the ATC is malfunctioning,
you can do without it in an emergency. For example, on the RCA TR -70,
simply pulling the Cavec module out from its back connector will open
the ATC error -correction circuitry. If Cavec is not used, the jumper on
pins 28 and 29 of the back connector of the Cavec position may be re-
moved to delete the ATC function. (Check your particular system for a
similar way of removing ATC.) Everything (headwheel, guide servo,
playback equalizers, etc.) must be adjusted perfectly for color operation,
but a program can sometimes be saved by this emergency procedure.
Keep in mind the three fundamental headwheel (drum) servo func-
tions and the basic ways to recognize their performance:
1. There is some form of voltage -controlled oscillator (vco) which
works from the comparison (frequency discrimination) of the head
tachometer with a fixed reference. This can be termed a time -
constant control.
If you place the picture monitor on external sync and play a tape
in a nonsynchronous mode, you would adjust any time -constant con-
trol for minimum horizontal roll on the monitor. This is strictly a
frequency control at the horizontal rate.
2. Velocity loop gain contains high -frequency positional errors. The
gain should just be sufficient to get as fast a lock-up as possible, but
not so "stiff" that there is instability on start -up, splice, etc. If the
gain is too low, the response is "'rubbery" after a transient.
3. Phase loop gain contains low- frequency errors for proper damping
of the velocity loop. The headwheel (drum) has a natural tendency
to "hunt" at between 10 and 30 Hz. At too low a gain, there will
tend to be a low- frequency "oscillation" in the drum error servo
outputs.
Servo adjustments are generally made with a condition of maximum
load. This simply means that the guide is placed in about a 1 -mil "heavy"
penetration while such adjustments are made. Then normal penetration
should result in tight control.
12 -9. BASIC APPLICATIONS AND

SERVICING OF LOGIC CIRCUITRY
Integrated circuitry and solid -state logic in general are fast replacing
discrete components and more cumbersome (older) circuit philosophy.
For example, the Ampex AVR -1 servo systems are 100 -percent logic, the
same as or similar to those described in Chapter 13. The latest cassette and
cartridge video-tape systems (Chapter 13) contain many such advance-
ments in video, fm, control, and servo functions. The principles of opera-
tion are identical to older ideas, but the results are obtained on a much
faster and more stable time base. We will now illustrate an application
of logic circuitry to a fast servo lock-up auxiliary development for the
RCA TR -70 known as Fablock. Chapter 13 will explore further the appli-
cations to cassette and cartridge video -tape systems.
NOTE: It is imperative that the reader have a background in logic cir-

cuitry at least equivalent to that of Chapter 3 of Television Broadcasting:
Systems Maintenance by Harold E. Ennes (Indianapolis: Howard W.
Sams & Co., Inc., 1972) .
The RCA Fablock

Fablock operates on two main areas of the machine, the take-up reel
motor boost control circuit and the circuits that limit the capstan servo
bandwidth (see Fig. 12 -22) One of the main causes of slow lock -up is
.
application of boost torque to the take -up reel for an unnecessarily long
period. Fablock overcomes this by monitoring the frequency of the con-
trol -track playback signal and removing the boost as soon as the tape
Processed TPI TP3 TP7

Control - o
Track
3 ms
Pulses IIP
r-f
H
2 ms Delay One -Shot
Interface Q4 Boost Control
111
Pulse Gen
Q1
UU3
Clock
Pos NAND OIP

Gate Clear Q-.-- Interlaces Servo Enable
U6 QS
Type -D
Flip -Flop
TP6
US IAI
US IBI
Delayed TP4
Type -D
Ref
Square 12 ms
Flip -Flop
Wave One Shot
D Set oTPB
Pulse Gen
U4
TP5
TP2 o
Tape
Frame 4.5 ms Invert OIP
Pulse One -Shot
Pulse Gen
Clock Level Change Interface --f
Q6 Q7 S rvo
U2
Co trol
Fig. 12 -22. Block diagram, Fablock main circuit board.
being played back reaches a speed corresponding to some predetermined

control-track frequency (typically 210 Hz) To allow the tape to reach .
this speed as quickly as possible, Fablock causes the capstan motor to be

driven at a constant rate, until the control -track signal exceeds the pre-
determined frequency, and then switches control of the capstan speed to
the capstan-servo error circuits.
An additional cause of slow lock -up is the relatively narrow bandwidth
normally used in the capstan servo. Because of the necessarily low sample
rate of the capstan servo, this bandwidth must be kept narrow during nor-
mal operation to prevent hunting. Fablock, therefore, improves the speed
of lock-up by introducing a new mode of capstan -servo operation referred
to as the wideband pull -in mode. In this mode, the Fablock temporarily
disconnects certain bandwidth -restricting components from the capstan -
servo error amplifier circuits. As a result, if the capstan servo measures
a large phase error during this period, the rephasing is carried out at a
rapid rate.
Entrance into the wideband mode is controlled automatically by a cir-
cuit called the near -lock detector, which is enabled as soon as the control -
track monitoring portion of the Fablock determines that the tape has
come up to speed. This circuit compares each successive tape frame pulse
with the corresponding delayed reference square wave and decides whether
or not the timed edges of these signals are within a predetermined range
(approximatly ±6 milliseconds) of coincidence. Approximate coincidence
occurs only when the capstan servo is nearly locked, since the capstan -
servo sample pulse is derived from the tape frame pulse and the capstan
trapezoid is derived from the reference square wave (Fig. 12-20B.)
If the circuit determines that near -lock does not exist, it provides a
control signal which places the capstan servo in the wideband mode until
the next comparison is made. As soon as near -lock is detected, however,
the circuit reinserts the bandwidth -restricting components, thereby allow-
ing the capstan servo to function normally.
Action of the Fablock is not restricted to the initial lock -up period after
entering the play mode. If an asynchronous switch occurs at any time
during playback, the machine temporarily reverts to the wideband mode.
During this period, the large capacitors in the bandwidth-determining
circuits, which have been temporarily disconnected, retain a memory of
the previously determined capstan error voltage. Therefore, when near-
lock is achieved and these components are reinserted, the long -term mem-
ory voltage is added to the short-term error voltage already present in the
servo error amplifiers. As a result, the capstan motor quickly reverts to
normal speed without overshooting excessively.
The main Fablock board includes two circuits, the control -track presence
and speed detector (portion shown above the dash line in Fig. 12 -22) and
the capstan servo near -lock detector.
Control -Track Detector In Fablock

The control -track detector (Figs. 12 -22 and 12 -23) includes inverter/
interface amplifier Q1, monostable circuits U1 and U3, positive -logic
NAND gate U6, Type -D flip -flop USA, and control transistors Q4 and Q5.
The input to the circuit consists of 10 -volt negative -going control -track
playback pulses (240/250 Hz) from the control -track record /play module
of the machine. These pulses are differentiated and applied to interface
inverter Q1. The output of Q1 is a positive-going pulse train going from
the low logic level of -5 volts to the high logic level of 0 volts. These
pulses are applied to the high -level actuated input (pin 5) of mono-
stable U1.
The period of U1, which is determined by R7 and C3 (Fig. 12 -23) is
approximately 2 milliseconds. Whenever the input goes high, therefore,
the Q output goes high and the Q output goes low for approximately 2
8ofr fr
oa _;..-t ó°a m N
tó fr
"-
O ¢S
g£ <
é ^ á
9
-
g
.-II"
t w
xI3 K1<g
g
c
:
; x '&1_
'--F---1--
.--11___7._"
B $
á
8g
á u Q S_ pS
ii
x
60- t
Oa O
o-
Z ó
00
y p
^ +S 5
Q
8 S
s
o-... ó
2f
o
_ S - -
/
Lo .. o , ---_
7-...
.
o
,,,
n Ñ 5 w. I( vn
., E
-0- ...á
E
4i:L<-:
S =
-¡(/ ^'
o
*5.
ry _
E
p R
..- II
-
3- 5-a
¢
'I"
vll
Fig. 12 -23. Schematic diagram, Fablock main circuit board.

milliseconds. The Q output of U1 is connected to the high -level actuated

input of monostable U3. As a result, U3 is triggered at the end of the
period of U1.
Monostable U3 has an unstable period of approximately 3 milliseconds,
which is adjustable by R15. The Q output of U3, therefore, which is con-
nected to the D input of flip -flop U5, goes high approximately 2 milli-
seconds after each control -track pulse and returns to low 5 milliseconds
(sum of periods of U1 and U3) after each control -track pulse.
The Q output of U1 is connected to the clock input of flip-flop U5.
The clock input, therefore, goes high whenever U1 is triggered. If the
period between control -track pulses is greater than 5 milliseconds (indi-
cating that the frequency of the control -track signal is less than approxi-
mately 210 Hz), the D input of U5A will already have returned to the
low level when each clock pulse occurs. As a result, U5A will be triggered
into (or remain in) the reset state (Q output low and Q output high) .
If, however, the period of the control -track signal is less than 5 millisec-
onds (indicating that the frequency is greater than approximately 210
Hz), the D input of U5A will still be high when the clock pulse occurs.
As a result, U5A will assume the set state. The exact control -track fre-
quency at which the changeover occurs may be adjusted with R15.
In addition to the D and clock inputs, U5A is controlled by the level
applied to the clear input from the junction of R18 and C11 in the output
circuit of NAND gate U6. The input to this gate consists of the Q outputs
of Ui and U3. When a control -track signal is present, these two outputs
are high simultaneously only for a period of approximately 30 nano-
seconds, which occurs while U1 is triggering U3. Since, however, R18 and
Cli provide a delay of approximately 100 nanoseconds, the clear input of
U5A never goes low, and the operation of the flip -flop is unaffected by
the clear input.
If, however, a satisfactory control track is not present, the Q outputs
remain high for longer than 100 nanoseconds. As a result, the clear input
goes low, thereby placing USA in the reset state.
The Q output of U5A is fed through a time -delay circuit consisting of
fixed resistor R17, rheostat R28, and capacitor C8 to interface /inverter
amplifier Q4. The Q output is fed to interface /inverter Q5 and to the set
input of flip -flop U5B in the near -lock detector circuit. The output of
Q4 (boost control) is fed to the take -up reel motor boost control circuit
on the control (or driver control) module of the machine. The output of
Q5 servo enable) is fed to the splice control relay in the capstan error
(
or capstan phase module.

When the machine is switched into the play mode, the Q output of
U5A remains low and the Q output remains high until a control -track
signal of the required frequency is present. Under these conditions, Q5
becomes saturated, thereby effectively grounding the servo enable bus. As
a result, the capstan servo will be disabled in a manner similar to that in
the record mode. The capstan motor will, therefore, rotate at the nominally
correct speed at this time and will not be affected by the capstan -servo
error. In addition, Q4 will remain cut off, thereby open- circuiting the
boost control bus and allowing normal application of boost to the take-up
reel motor.
The low level from the Q output of USA is also applied to the set input
of U5B in the near -lock detector ( described later) . As a result, U5B is
held in the set state at this time, thereby disabling the near -lock detector.
As soon as the control track reaches the required speed, U5A becomes
set, as previously described. The Q output therefore goes high and enables
the near -lock detector. In addition, Q5 becomes cut off, thereby enabling
the capstan servo.
The Q output of USA also goes low at the same time, but Q4 still re-
mains cut off for a delay period which is adjustable by R28. At the end
of this delay, Q4 becomes saturated and disables the boost control circuit.
The delay period is required because the slack tape, which is shared
equally by the two tension arms on the transport at the instant of starting,
is all collected at the arm nearer the take -up reel by the time the control -
track detector registers the up -to -speed condition. Therefore, to prevent
this arm from bottoming and actuating the tape-break switch, boost must
still be applied for a short additional period.
Capstan Servo Near -Lock Detector In Fablock

The near -lock detector (Figs. 12 -22 and 12 -23) consists of inverter/
interface amplifiers Q2 and Q3, one-shots U2 and U4, type -D flip -flop
USB, inverter /level-change amplifier Q6, and transistor switch Q7. The
inputs to the circuit consist of the delayed reference square wave and the
tape frame pulse from the capstan servo. In the capstan servo, the delayed
reference square wave is used to generate the capstan trapezoid, and the
tape frame pulse is used to generate the sample pulse, which becomes
locked to the trapezoid waveform when the servo has stabilized. Thus,
when the reference square wave and tape frame pulse are almost coinci-
dent, the capstan servo will be nearly locked.
The reference square wave, which has a frequency of 30 or 25 Hz (15
or 12.5 Hz during fast lock -up capstan interval) , is differentiated and in-
verted by Q3 to form positive -going trigger pulses between -5 and 0
volts. These pulses are fed to the high -level actuated input of U4. The
unstable period of U4, which is adjustable by bracket -pulse potentiometer
R16, is approximately 12 milliseconds. The Q output of U4 is fed to the
D input of flip-flop USB. Thus the D input goes high for approximately
12 milliseconds each time that U4 is triggered.
The tape frame pulses are differentiated and inverted by Q2, to form
positive -going 5 -volt trigger pulses which are fed to U2. The unstable
period of U2, which is adjustable by the sample-pulse delay potentiometer,
R9, is approximately 4.5 milliseconds. The Q output of U2 thus goes low
when U2 is triggered, and it returns to the high state 4.5 milliseconds

later. This output signal is fed to the clock input ( terminal 11) of flip-
flop U5B.
As previously mentioned, before the control track has come up to speed,
a low level is fed from the Q output of U5A in the control -track detector
to the set input of U5B, thereby holding U5B in the set state (Q output
high). As soon as this low is removed, however, control of U5B is trans-
ferred to its D and clock inputs. The clock pulse occurs 4.5 milliseconds
after U2 is triggered by the tape frame pulse, and the D input is high for
12 milliseconds after U4 is triggered by the reference square wave. Thus,
if the timed edges of the tape frame pulse and reference square wave are
more than approximately --6 milliseconds apart, the D input will be low
when the clock pulse occurs, and U5B will, therefore, be triggered into
the reset state. When, however, the timed edges are within ±6 milliseconds
of each other (defined as the near -lock condition) , the D input will be
high when the clock pulse occurs. As a result, U5B will be triggered into
the set state and will remain in that state as long as near -lock exists. As
previously mentioned, the timing between the outputs of U2 and U4 de-
pends on adjustment of the sample -pulse delay potentiometer, R9, and
the bracket -pulse potentiometer, R16. These potentiometers are normally
adjusted so that when the capstan servo is locked, the trailing (positive -
going) edge of the pulse from the Q output of U2 occurs in the center
of the pulse period of U4.
The Q output of U5B is fed to inverting and level-shifting amplifier
Q6, which controls transistor switch Q7. When U5B is in the set state,
the Q output is high (at 0 volts), and Q6 is saturated. As a result, Q7
conducts and grounds resistor R29. Since this resistor, in series with po-
tentiometer R30 (servo time -constant control) , is connected to the damp-
ing components of the capstan-servo error amplifier, these components
are inserted in the circuit, thereby allowing the capstan error circuits to
function normally. When, however, U5B is in the reset state, the Q out-
put is low. As a result, both Q6 and Q7 are cut off, thereby disconnecting
the damping components.
From the preceding, it can be seen that when U5B is initially held in
the set state before the control track has reached the desired frequency,
the damping components are in the circuit, thereby allowing a memory
voltage to build up on the damping capacitors. After the control track
has reached the desired frequency, U5B is triggered into the reset state by
the next clock pulse, thereby removing the damping components and
speeding up the capstan servo. Then, when near -lock is detected and the
damping components are reinserted, the previously developed memory
voltage is reapplied, thereby preventing the servo from overshooting ex-
cessively. The servo time -constant control, R30, permits adjusting the
damping characteristics. This control is normally adjusted for critical
damping as described under setup adjustments.
Fablock -5 Volt Regulator

A power- supply circuit consisting of transistor Q8, zener diode X3,
resistors R25 and R26, and capacitors C9 and C10 is provided on the
Fablock board. This circuit reduces the -10 -volt input from the machine
to a regulated -5 volts as required by the Fablock logic circuits.
Fablock Subboards A and B

The circuits of Fablock subboards A and B are shown in Fig. 12 -24.
Only one of these boards is actually used in a particular installation, de-
pending on the type of machine. Subboard A is installed in the control -
track record /play module of the TR -50, TR-60, TR -70, TR -70A, TR -70B,
TR -22D, TR -3, TR-3A, TR -3B, TR -4A, TR -4B, and TR -4C machines.
Subboard B is installed in the corresponding module of the TR -22A, TR-
22B, TR-22C, and TR -4 machines. The boards perform identical functions
but differ slightly to suit the requirements of each control -track circuit.
These boards are required because the control -track playback amplifier
in its original form is a 240- or 250 -Hz tuned amplifier with so much gain
that it provides an appreciable 240- or 250 -Hz output due to amplification
of cross talk and noise, even when no control track is present. Although
this spurious signal does not affect normal non -Fablock operation, it is
undesirable since it could cause the Fablock circuit to register the presence
of a satisfactory control track when none is present.
Subboard A or B is inserted at a point in the amplifier where the am-
plitude of the cross talk and noise is considerably lower than the level of
a normal control -track signal. The circuit on the board provides a coring
action which causes the amplifiers following it to ignore all signals below
a level greater than that of the cross talk and noise. Thus, when no control
track is present, the control -track playback amplifier output is zero.
Fablock Setup Procedure
We will describe the rather simple setup procedures for this circuitry
so that a brief outline of maintenance and servicing techniques can be
included.
xi X
(A) Subboard A. (B) Subboard B.

Fig. 12 -24. Schematic diagrams, Fablock subboards.
Control -Track Sensor Adjustments -

1. Make a stable recording of 5 minutes duration.
2. Select the CAP ERROR display on the cro.
3. Enter the standby mode and, while depressing the CAP SET button on
the capstan oscillator module, adjust the capstan free -running speed
to give minimum drift between the sine -wave and spike displays.
4. Select the tonewheel playback mode, and switch the picture monitor
to external sync.
5. Maintaining pressure on the CAP SET button, enter the play mode,
and observe the picture monitor for overspeed characteristics. If R15
in the Fablock circuit is correctly adjusted, the tape should not over -
speed. Repeat this step after adjustment of R15, if necessary.
6. Wind to the end of a 90- minute tape, and make a stable recording
of 5 minutes duration. Make adjustments to R28, as required, so that
on starting, boost is maintained for a sufficient time to avoid build -up
of a loop at the take -up damping arm. If the range of R28 is insuffi-
cient, the boost voltage should be moved to the next higher trans-
former tap.
7. Remove the control -track record /play module, and enter the play
mode. Make a final adjustment of the CAP FREQ control to obtain
minimum run -through as observed on the fm position of the cro.
Near -Lock Detector Adjustments -
1. Turn R9 (sample -pulse delay) and R16 (bracket -pulse width) on
the Fablock unit to the maximum clockwise position.
2. Enter the play mode, and wait for the machine to lock. With a dual -
trace oscilloscope, observe TP2 and TP4 on the Fablock unit.
3. Adjust R16 to give a positive pulse width of 12 milliseconds at TP4.
4. Adjust R9 to bring the trailing edge of the narrow, negative -going
pulse on TP2 to midway through the positive period of the wave-
form on TP4.
5. Stop the machine, enter standby, and re -enter the play mode. The
machine should then lock within the specified time of two seconds.
If performance is unsatisfactory, adjustment of R30 is required.
Potentiometer R30 in the Fablock unit is associated with the damp-
ing capacitor in the basic servo error amplifier. It should be adjusted
to give minimum overshoot as seen on the capstan trapezoid display.
NOTE: On either side of the correct setting of R30, excessive hunting

will be experienced. Owing to individual variations between machines,
fine adjustments to R16 and, consequently, R9 may be found to offer an
improvement in lock -up time. In general, however, R16 should not be
adjusted to give more than one half -cycle of overshoot before lock, to
avoid a permanent hunting condition. (This condition will occur if the
near -lock range is made too narrow.)
Maintenance Procedure for Fablock

An outline of the maintenance procedure for this circuitry will serve to
give the reader a basic technique for troubleshooting of logic circuitry in
general.
The precautions normally taken when working with semiconductors and
printed circuits should be observed when working on the Fablock board.
No more heat than necessary should be applied when removing suspected
parts, to avoid damage to both components and the bond between the
board and the printed wiring. In general, restrict the application of heat
to a maximum of 5 seconds at a time. Use of a pneumatic or other form
of desoldering aid is recommended. After a replacement of any transistor,
ensure that the plastic spacer is retained in position. This will afford in-
creased thermal protection when resoldering.
To aid servicing and replacement, all integrated circuits are plugged
into sockets. They should be removed from the sockets before any solder-
ing is attempted on these sockets. Since semiconductors, in general, are
sensitive to any voltage overload, it is advisable to remove power before
any maintenance is attempted.
The following equipment will be required for troubleshooting the
Fablock.
1. Tektronix Type 545 oscilloscope with dual -trace plug -in, or equiv-
alent.
2. Test meter with 20,000 ohms /volt sensitivity and ranges suitable
for checking 5, 10, and 70 volts dc.
3. Normal service tools, including:
(a) Soldering iron (25 -watt)
(b) Desoldering tool
(c) Long -nosed pliers
(d) Side-cutters
(e) Screwdrivers
This section should be used in conjunction with the preceding discus-
sion of the theory of operation. Before starting any fault- finding routine,
check that the derived power- supply voltage of -5 volts is present and
within its tolerance of ±0.25 volt.
The signal path through the Fablock unit, including test points, is
shown in the simplified block diagram of Fig. 12 -22. If a fault should
develop in operation, this diagram should be consulted to determine which
inputs are required by the nonfunctioning area, and a logical progression
of checks should be carried out from input and output. The following
indicates a suggested troubleshooting procedure.
1. Capstan servo remains in record mode, evidenced by sine -wave and

spike display on monitoring cro during playback.
(a) Switching transistor Q5 (servo enable) could be short-circuited.

Check Q5.
(b) Fault in Ul, U3, U5, or U6 (associated with control-track speed
detection). Check U1, U3, U5, and U6.
(c) No feed of processed control -track pulses to Fablock. Check
for presence of control-track pulses at pin 1.
(d) Control -track control R15 misadjusted. Check adjustment of
R15 as directed under Setup Procedure.
2. Monitoring cro display switches rapidly between trapezoid and sine
wave during servo lock-up.
(a) Control -track detector set at too high a frequency. Turn con-
trol -track potentiometer (R15) clockwise.
3. Trapezoid display continuous in playback, but servo hunts rapidly
and locks up only after a considerable time, if at all.
(a) Sampling delay off center, or bracket -pulse setting too narrow.
(b) Capstan damping incorrect. Increase settings of bracket -pulse
control (R16) and sample -pulse delay control (R9) to maxi-
mum clockwise, and recheck operation. If the operation is now
satisfactory (hunting eliminated) , carry out final setup in ac-
cordance with Setup Procedure section. If it is still unsatisfactory,
set R16 to give a 12- millisecond pulse at TP4, set R9 to place
the positive edge of the TP2 waveform at the midpoint of the
TP4 positive pulse, and adjust R30 for minimum hunting. With
R30 too far counterclockwise, a slow hunting condition will
occur; if it is too far clockwise, the hunting will be more
rapid. A critically damped condition can be reached between
these two extremes; this condition results in a much reduced
hunting period.
If the operation is still unsatisfactory, proceed as follows:
1. Withdraw the capstan error module (TR -22 and TR -70 series) or
capstan oscillator module (TR -3, TR -4, TR-50, TR -60 series).
2. Locate the C11 positive terminal (TR-22, TR-22A, TR-22B, TR-
22C) or the C17 positive terminal (TR -22D, TR -70, TR-70A,
TR -70B) or the C39 positive terminal (TR -3, TR -4, TR -50, TR-
60 series) Remove the Fablock lead from this terminal and con-
.
nect an 1800 -ohm resistor to ground.

3. Reconnect the module to the machine either by reinserting it di-
rectly or, preferably, in a module extender, and initiate the play
mode. The machine should now lock up in identical fashion to a
machine not equipped with the Fablock accessory. If it does not,
a more basic servo fault is indicated, and the servo maintenance
manual should be consulted. If the machine does lock, with the
machine in playback carry out the checks in the following steps.
4. Pin 5, TP8, should be at a level between -2 volts and 0 volts (as

opposed to -3 to -5 volts). (This will be true only if the tape be-
ing played has a satisfactory control track.)
5. Pin 12, TP4, should have a positive -going pulse of approximately
12 milliseconds duration and greater than 2.5 volts amplitude, sit-
ting on a base line between -5 and -4.5 volts, and occurring at
the frame rate (once every two frames during fast lock -up capstan
interval on either 525/15-Hz edit pulses or 625 -line high -band op-
eration) Absence of these pulses indicates a malfunction in the
.
area of U4, or absence of delayed reference square -wave input.

6. Pin 11, TP2, should have negative -going pulses of approximately
5 milliseconds duration and greater than 2.5 volts amplitude with
negative peaks between -5.0 and -4.5 volts, and occurring at
the frame rate (once every two frames during fast lock-up capstan
interval on either 525/15 -Hz edit pulse or 625 -line high -band
operation) Absence of these pulses indicates a malfunction in the
.
area of U2, or absence of input tape frame pulses.

7. Confirm that the positive -going edge of the pulse at TP2 occurs
during the positive excursion of the pulse at TP4. (This requires
the use of either a dual -trace oscilloscope or a single -trace oscillo-
scope with facilities for external, delayed triggering from reference
vertical) Ideally, the positive -going edge of the pulse at TP2
.
should occur halfway through the duration of the pulse at TP4, and
it should be so adjusted with the sample -pulse delay control (R9
in Fig. 12 -23) .
8. A satisfactory checkout of all conditions in steps 5, 6, and 7 should

give rise to a logic high condition at pin 9 of U5 (0 to -2 volts,
as opposed to -3 to -5 volts). Failure to obtain a logic high at
pin 9 with all inputs correct points to a fault condition in U5. In
the event of a satisfactory logic high and continuing faulty opera-
tion, proceed to the next step.
9. Check the operation of Q6 and Q7. With a logic high at pin 9 of
U5, Q6 should saturate, and the Q6 collector should be at approxi-
mately -4.5 volts. This causes Q7 to saturate, which in turn grounds
R29 in Fablock, resulting in normal, reduced servo bandwidth.
10. Check for continuity between Fablock Pin 7 and the C11 positive
terminal in the capstan error module (TR-22, TR -22A, TR -22B,
TR -22C), the C17 positive terminal in the capstan error module
(TR -22D, TR-70, TR-70A, TR -70B), or the C39 positive terminal
in the capstan oscillator module (TR -3, TR -4, TR-50, TR -60).
12 -10. AUDIO SYSTEM MAINTENANCE
The following major pieces of test equipment are required for audio
system checks and adjustments:
1. Audio signal generator (Hewlett -Packard Model 206A or equivalent)

2. AC voltmeter (Hewlett -Packard Model 400 or equivalent)
3. Vacuum -tube voltmeter or sensitive volt-ohmmeter
4. Oscilloscope (Tektronix Type 547 or equivalent)
5. Distortion meter ( Hewlett- Packard Model 330C or 331A, General
Radio Type 1932A, or equivalent)
6. Degausser (supplied with machine)
7. Test module (supplied with machine)
The audio heads may eventually become slightly magnetized. If the

magnetism is not removed, it could cause distortion, loss of high- frequency
response, and in particular, increased tape noise in the output. However,
the heads may easily be demagnetized with the degausser provided in the
accessory kit included with the machine.
For best signal -to -noise ratio, each audio head should be degaussed be-
fore every critical recording session as follows:
1.Place the EQUIPMENT POWER circuit breaker in the off position, but
leave the UTILITY POWER circuit breaker on so that power is supplied
to the convenience outlet.
2. Connect the degausser power cord to the convenience outlet on the
front of the recorder.
3. Place the tip of the degausser gently against the pole pieces of the
head, and slowly move the degausser from side to side two or three
times with the degausser just barely in contact with the head poles.
4. With the degausser poles directly over the head poles, lift the de-
gausser off the head, and withdraw it very slowly and evenly. Then
unplug the degausser power cord.
CAUTION: To avoid scratching the heads, never allow metal objects to
contact the head posts or the pole pieces of the heads. If the plastic coat-
ing on the tip of the degausser wears off, place masking tape on the de-
gausser before using it.
Audio-system tests and adjustments are facilitated by the fact that the
audio simulplay head makes it possible to adjust the audio recording
system while simultaneously checking the results through the playback
and monitoring systems. The purpose of these adjustments (all of which
occur in the recording system) is to permit making recordings that pro-
duce the specified signal -to -noise ratio and to provide a flat overall record/
playback frequency response through the audio range.
To illustrate general audio -system maintenance techniques for the video
tape recorder, we will present the procedures recommended by RCA for
the TR -70. Since the adjustments depend on each other, they should be
performed in the sequence presented. After the adjustments are com-
pleted, the signal -to -noise ratio and the overall record /playback frequency
response should be checked.
Test Setup
The audio - system tests are performed with a tape speed of 15 inches
per second except where noted otherwise. All audio covers and shields
should be in their proper places. The audio monitor amplifier is con-
nected for monitoring. Fig. 12 -25 shows the test -setup connections. These
connections are changed as directed in the procedures, in order to accom-
plish the specific test and adjustment in progress.
Front Rear
TP2 12J21
Vacuum
Audio Null r Audio In Audio Signal
Tube Ai 7 Generator
Volt meter
TP3 b } o
B
Q i
i
Cue Null 12.122 /
Audio Record
Amplifier
14011
Cue Record
Amplifier
14031
..." 12124
Cue In
AC
Voltmeter
1_7
s
T P4
Audio Eraseó
f'
Erase
M011
-
Bias
Audio Cue
Preamp
14171
'X
12125
Audio Cue Playback Cue Out
Playback Distortion
Oscilloscope Amplifier
Amplifier Meter
TP5 14341
14311 ,
Cue Erase J
AC
12.123
Voltmeter
TR -70 Audio System
Audio Out
Fig. 12 -25. Test setup for RCA TR -70 audio- system checks and adjustments.
Audio Test
The audio test requires that the machine be in normal operating con-
dition, with tape in the machine. Proceed as follows:
1. Connect the audio signal generator to the AUDIO LINE A IN jack
(12J21). The generator is connected across pins 2 and 3 of the jack.
Pin 1 is ground. Then switch to Video A.
2. Set the audio signal generator for a balanced output of 600 ohms.
Terminate the generator in 600 ohms.
3. Connect the distortion meter with a 600 -ohm load resistor to the
AUDIO LINE OUT jack (12J23). Pins 2 and 3 are across the output.
Pin 1 is ground.
4. Set the LEVEL control on the record amplifier module for maximum
gain (fully clockwise) .
5. Adjust the audio signal generator for a 1000 -Hz output. Adjust the
output level to -10 dBm.
6. Turn the E-E switch (on modulator afc module 509) on. The input
signal should now be present at the output and at all audio monitor
positions.
Erase -Head Tuning and Noise -Null Adjustments
In the audio -system tests and adjustments, the head -tuning and noise -null
adjustments are accomplished as follows:
1. If the AUDIO BIAS control (R33) , on the audio /cue preamplifier

module (module 417) , has been disturbed or replaced, preset it to
a two -thirds position in the clockwise direction.
NOTE: The adjustment shaft of control R33 is located on the upper side
of module 417 and slightly right from center. The module is located on
the tape transport, just above the heads.
2. If the audio record high- frequency equalization trimmer capacitor

(C16) has been disturbed or replaced, preset it to one -quarter turn
from fully tightened (counterclockwise) .
NOTE: Trimmer capacitor C16 is located slightly right from center.

3. Connect a dc voltmeter between the AUDIO NULL test point and
ground on erase oscillator module 601.
4. Adjust the AUDIO DC NULL control, on the erase oscillator module,
for a zero indication on the voltmeter.
5. Place the EQUIPMENT POWER circuit breaker in the off position,
and carefully degauss all heads.
6. Place the EQUIPMENT POWER circuit breaker in the on position.
7. Terminate the AUDIO IN (12J21) line with 600 ohms (or switch
in full attenuation on the audio signal generator).
8. Place the machine in the setup mode.
9. Connect the oscilloscope to the AUDIO ERASE test point (TP4) on
the erase oscillator module and look for 140 volts peak to peak. If
the voltage level is not at 140 volts, place the erase oscillator module
on an extender frame and adjust the audio erase level control (R96)
for the required indication of 140 volts peak to peak.
10. Remove the preamplifier shield cover, and connect the oscilloscope
to the audio record input point on audio /cue preamplifier 417.
This is at capacitor C17, which is located on the lower side of the
board, to the extreme right.
11. Adjust bias trap coil L1 for a minimum signal indication, which
should be below four volts peak -to -peak. The adjustment screw of
coil Ll is located adjacent to control R33.
12. Remove the oscilloscope leads.
13. Press the AUDIO ON push button on the record panel, and make a
recording with no input signal.
14. Press the AUDIO PB push button on the audio switcher. Rotate both
the LEVEL control on audio playback module 431 and the audio
monitor SPEAKER VOLUME control to fully on (clockwise) while
listening to the simultaneous playback of the recording.
15. Adjust the AUDIO TUNE control (C31) on erase oscillator module
601 for minimum noise level as heard on the loudspeaker.
NOTE: Noise indications may be observed on the ac voltmeter (from

12J23) as sharp fluctuations over a steady indication of approximately
0.08 volt, which is caused by the bias signal. Also, the meter input termi-
nals may be shunted with a 0.1 -p,F capacitor to permit noise indications
to be observed directly on the ac voltmeter.
16. Adjust the AUDIO DC NULL control (R88) on the erase oscillator
module for further noise reduction. This adjustment should be
performed carefully to obtain the lowest tape noise. Rotate the
control no further than necessary in either direction of minimum
level indication to find the null point. The noise level is usually
0.04 volt with the machine stopped, or 0.05 volt with the machine
operating in the play mode over this recording. Indications are ob-
tained with the ac voltmeter.
Bias and High- Frequency Control Adjustment
Proceed as follows to adjust the bias and high- frequency controls:
1. Record a 1000 -Hz signal with a zero -level indication on the VU

meter while the AUDIO REC push button on the audio switcher is
depressed.
2. Press the AUDIO PB push button on the audio switcher, and adjust
the LEVEL control on audio playback module 431 to obtain an indi-
cation of approximately -3 on the VU meter.
3. Adjust the AUDIO BIAS control (R33) on the audio /cue preamplifier
module (module 417) for maximum signal-output indication on the
VU meter. If the VU -meter indication exceeds zero level, reduce the
LEVEL -control setting on the audio playback module (module 431)
so as to obtain a true peak setting for the AUDIO BIAS control.
4. Measure the bias voltage on the audio record /play head (2PU3) ,
from pin 3 of the head to ground. Record this value.
5. While recording a 1 -kHz signal, adjust the playback LEVEL control
for a 0 indication on the VU meter.
6. Set the signal -generator frequency to 15 kHz, and hold the generator
output level the same as at 1 kHz in step 5 above.
7. Adjust the audio record high- frequency equalization control (C16)
for the same output level as noted in step 5.
8. Set the AUDIO BIAS control (R33) on module 417 to maintain a

constant bias voltage as determined in step 4.
9. Repeat steps 5, 6, and 7 if necessary.
Distortion Test
The distortion test and associated adjustments are accomplished as
follows:
1. Connect the distortion meter in parallel with the voltmeter that is

across the output of the machine.
2. Adjust the audio signal generator for a 400 -Hz output, and set
the level to 10 dBm (2.45 volts) Set the attenuator to 10 dB.
.
3. Rotate the LEVEL control on audio record module 401 to its maxi-
mum clockwise position.
4. Record a one-minute section of tape at this input level; then de-
crease the attenuation in 1 -dB steps to +7 dBm while recording
each level for approximately one minute. While recording, position
the MIC switch to Audio, and announce the input level into the
microphone; then return the MIC switch to the off position.
5. Rewind this section of the tape and play it back. Set the LEVEL
control on module 431 for approximately 2.0 volts, as indicated on
the output ac voltmeter.
6. Place the distortion meter in the calibrate mode, and adjust the
distortion -meter input control for a full-scale meter indication; then
read the overall distortion.
7. Reset the input control on the distortion meter as the level changes,
according to the level announced in step 4, and read the overall dis-
tortion.
8. Determine what input level is required for the distortion to reach
3 percent (if distortion is excessive, begin recording at a lower
level).
9. Reset the attenuator to the level determined in step 8, and place
10 dB of additional attenuation in the audio generator.
11. Press the AUDIO REC push button on the audio switcher.
12. Mount the audio record module (module 401) on an extender
frame, and adjust the record -level meter calibration control (R47) ,
located near the rear half of the module, to obtain a zero indication
on the VU meter at this input level (10 dB below the input for
3-percent distortion).
13. Place the machine in the stop mode.

14. Turn on the E-E switch on the 509 module panel.
15. Press the AUDIO PB button on the audio switcher.
16. Adjust the playback-loop level control (R45) for zero indication
on the VU meter, with the same input as in step 12.
17. Remove the module extender frame, and replace the module in its
proper place in the machine.
18. Record a section of tape at this 0 -VU recording level, and play it
back. The distortion should not exceed 1.5 percent.
19. Press the AUDIO REC button on the audio switcher. Reduce the
attenuation of the signal generator 10 dB; then press the vu -10
DB button on the audio switcher, and note that the meter indication
decreases to 0 (± 1) VU.
Signal -to -Noise Ratio Measurement

The signal -to -noise ratio is measured as follows:
1. Replace all audio shields, and record a 400 -Hz tone on a section of
tape at a 3- percent distortion level; then play it back.
2. Set the LEVEL control on the audio playback module (module 431)
for an indication of 3.2 volts (10 dBm) on the ac voltmeter.
3. Adjust the input sensitivity control on the distortion meter for 0 -dB
reference at the present level setting.
4. Rewind the tape to the starting point of the 3- percent distortion re-
cording, and terminate the audio input terminals (or apply full at-
tenuation to the audio signal generator) .
5. Operate the machine in the audio record mode over the section of
tape having the distortion recording to erase the tape with the audio
erase head only.
6. Rewind the tape to the beginning of this recording, and play the
tape back. Observe that the noise voltage level is at least 55 dB be-
low the output voltage at the 3- percent distortion level. The equiv-
alent voltages are a 0.005 -volt noise level with the 3.2 -volt setting.
7. Rewind the tape to a section where the audio is erased; then place
the machine in the play mode.
8. Operate the machine into and out of the audio record mode, and
note at what points on the counter the changes are made.
9. Play the tape back, and observe that the switching transients re-
corded are at least 40 dB below the 3- percent level, as indicated on
the distortion meter.
E -E Frequency Response
The playback loop is checked for frequency response as follows ( this
circuit may also be referred to as the back -to -back function or electrical -
to- electrical circuits) :
1. Set the tape speed to 15 inches per second.

2. Adjust the audio signal generator to supply 1000 Hz at 10 dBm,
and add an attenuation of 10 dB.
NOTE: Keep the output level of the audio signal generator constant.
3. Press the AUDIO PB button on the audio switcher.

4. Place the machine in the stop mode, and set the E -E switch on the
509 module to On.
5. Adjust the distortion -meter level control for a 0 -dB indication.
6. Measure the frequency response from 50 Hz to 15 kHz. The limits
should be plus zero or minus 1.5 dB.
7. Set the tape speed to 7.5 inches per second, and repeat the entire
procedure at this tape speed.
Record /Playback Frequency Response
The frequency response for tape speeds of 15 and 7.5 inches per second
in record and playback is tested as follows:
15 Inches per Second-

1. Set the tape speed to 15 inches per second.
2. Adjust the signal generator to supply 1000 Hz at 10 dBm, and add
an attenuator of 10 dB (0 dBm output level) .
NOTE: Readjust the level of the audio signal generator as required to
maintain a constant output level.
3. Press the AUDIO REC button on the audio switcher.
5. Adjust the LEVEL control on the audio record module (module
401) for an indication of zero on the VU meter.
6. Noting that the output meter of the audio signal generator is held
constant at 10 dBm, record the frequencies in Table 12 -4, starting
at 1000 Hz and disregarding changes of the audio record level as
indicated on the VU meter.
7. Switch the MIC AUDIO /CUE selector switch to the Audio position,
and announce each frequency in the microphone; then return the
switch to off while recording the frequency.
8. Terminate the output line with a 600 -ohm resistor.
9. Rewind and play back this recording, and adjust the LEVEL control
on the audio playback module (module 431) for 1.23 volts rms ( 4
dBm), as indicated on the ac voltmeter, while playing back the
1000 -Hz reference tone.
10. Press the AUDIO PB button, and note that the VU meter indicates 0
(±1) VU.
11. Adjust the input sensitivity control on the distortion meter for
zero dB, while the meter function is positioned for set level, and
the meter range is set for 100 -percent indication.
12. Measure the frequency response as indicated by the distortion
meter. For high- frequency response adjustments, if required, refer
Table 12 -4. Frequency- Response Chart
Audio Output (dB) Cue Output (dB)

Frequency
(Hz) 15 in /s 7.5 in /s 15 in /s 7.5 in /s
1000 0 (Ref) 0 (Ref) 0 (Ref) 0 (Ref)

50 ±2 - ±2 ±3
60 - ±2 - -
1 00 -I-2 -1-2 -
±2
-
±3
190
200 -1-2 -E2 - -
310 - ±2
- ±2
-
±3
-
400 -1-2
1000 ±0.5 i-0.5 +0.5 -*-0.5
3000 ±2 ±2
5000 ±2 ±2
7000 ±2 ±2 ±2 ±3
10,000 ±2 ±2 ±2 ±3
12,000 -1-2 - - -
15,000 -E2 - - -
to procedures given previously under Bias and High- Frequency
Control Adjustment.
7.5 Inches per Second-
1. Switch the tape operating speed to 7.5 inches per second, and repeat
the entire procedure at the reduced tape speed. The record LEVEL
control on module 401 is set at 1000 Hz for an indication of -10
VU for the 7.5 -in /s procedure.
2. Mount the audio record module (module 401) on an extender frame.
3. Operate the machine at 7.5 inches per second while operating in the
audio record mode.
4. Make a 1000 -Hz recording at the -10 VU record level, and adjust
the audio -output LEVEL control on module 431 for 0 VU output.
5. Set the audio signal generator to supply a 10 -kHz output at the same
oscillator output level. Disregard an increase in record VU -meter
indication.
6. If necessary, adjust the 7.5 -in /s high- frequency equalization control
(R32) , on the audio record module (module 401) , to obtain the
same playback output level at 10 kHz as at 1000 Hz. Use the dis-
tortion meter for an output indicator.
NOTE: Control R32 is located just forward of the center of the board and
on the upper half.
7. Rewind the tape to the starting point of this recording.

8. Play back the recording, and note that the response is the same for
1000 Hz as for 10 kHz (if not, readjust control R43 in audio record
module 401) .
EXERCISES
Q12 -1. Name two head faults not under your control, and describe how to
check for these faults.
Q12 -2. If you notice on a color monitor that 3 or 4 bands of 16 lines each
(indicating one of the four video heads) have excessive color beats
(moiré) resembling herringbone interference, what is the most prob-
able cause?
Q12 -3. What is the quickest way to check amplitude -vs- frequency response
of the E -E, playback, and record functions?
Q12 -4. Give the proper time base to use on the Tektronix Type 529 waveform
monitor to evaluate correctly (A) the T-pulse and (B) the 2T -pulse
on the cro graticule.
Q12 -5. What test signal most conveniently evaluates luminance-to- chromi-
nance delay inequality?
Q12 -6. What color -picture impairment results from luminance-to-chromi-
nance delay inequality?
Q12 -7. What factors other than electronic circuit problems can cause poor
servo stability and slow lock -up?
Q12 -8. What factors other than electronic circuit problems or misadjustments
can cause poor audio performance?
CHAPTER 13
Quadruplex -Tape
Cassette / Cartridge Systems
A system for the production of programming and station breaks from

a number of short prerecorded video tapes, generally through the use of
dual tape decks, is termed a cassette system by Ampex and a cartridge
system by RCA. The same quadruplex rotating headwheel is used in the
cassette /cartridge machine as is used in the larger systems previously
described.
The Ampex ACR -25 automatic video cassette recorder /reproducer sys-
tem may be purchased as a completely independent system or as a unit
which shares certain electronics with the Ampex AVR -1. Similarly, the
RCA TCR -100 can be obtained as a completely independent system, or as
one that shares electronics with an RCA TR-70.
13 -1. THE AMPEX ACR -25

The ACR -25 is a dual -transport, cassette- loaded, vacuum -threaded, two -
inch quadruplex video tape recorder /reproducer. It consists of two units,
the transport console (Fig. 13 -1A) and the control console (Fig. 13 -1B) .
The operator has random access to 24 cassettes which are contained in a
carrousel located between the two tape transports. Each cassette has a
capacity of 6 minutes, plus leader and trailer, at 15 in /s (12 minutes at
71/2 in /s). The machine records in high band only, but it will play back
high band, low band, monochrome, or color. Recordings made on a con-
ventional reel -to -reel machine can be spooled into cassettes and played back
on the ACR -25. Conversely, tapes recorded in the ACR-25 can be put on a
reel and played on a conventional vtr.
Performance specifications of the cassette machine are the same as those
of the Ampex AVR -1, due to the use of the AVR -1 signal and time -base-
correction systems. Ten -second spots can be played back -to -back, and lock-
up time is 0.2 second (0.35 second for 625 -line standards). By utilizing
451
the rewind lock -out feature on longer program segments, programs up to

six minutes in length can be played back-to -back with ten -second spots. In
this mode, the cassette does not rewind after play is complete, but is re-
turned immediately to the carrousel. It can then be rewound manually at a
later time when the ACR -25 is off line.
Left Transport Carrousel of Right Transport Automatic Monitoring Manual
(Al 24 Cassettes IBI Control Panel Deck Control Panel
I
i
Signal Control Panel Setup Control Panel

(Under Cover) (Under Cover)
(A) Transport console. (B) Control console.

Fig. 13 -1. Ampex ACR -25 cassette tape system.
As stated earlier, the ACR -25 records in high band only. The various
record modes are: (1) onto either the A or B transport, (2) onto both the
A and B transports simultaneously from the same program source, and
(3) B-to -A or A -to -B dubs within the machine. A mechanical record lock-
out switch is available on each cassette as protection against accidental
erasure of a master tape.
In addition to the standard equipment configuration, the ACR -25 can
also be used in conjunction with an AVR -1 to time -share the time -base-
correction electronics.
Since all of the AVR -1 design features are incorporated in the ACR -25,
accessories such as auto -chroma, velocity compensator, auto -tracking, drop-
out compensator, etc., are also available on the ACR -25. Two additional
QUADRUPLEX-TAPE CASSETTE /CARTRIDGE SYSTEMS 453
accessories which contribute to the back -to -back playback capability are:
(1) control -track rewrite, which allows the operator to record a new control
track on the tape if the original one is unusable, and (2 ) automatic play-
back adjust, which allows the operator to record on the cue track, in digital
form, information that will correct chroma phase, video gain, audio gain,
and black level on a tape that was not recorded correctly. This information
is recorded only once, and every time the tape is played thereafter, it
will be correct.
Many of the controls on the ACR -25 can be located at a remote point,
and the machine can be interfaced with a station computer.
Transport Console
The transport console contains two identical vacuum -threaded tape trans-
ports; a carrousel; an internal air system; head, capstan, and carrousel motor
drive amplifiers; transport power supply; main power transformer; cassette
carriage; and transport electronics bay. The cassettes are moved from the
carrousel to the transport loading area by the cassette carriage. The tape
is then withdrawn from the cassette and threaded into the transport by the
vacuum threading process. Fig. 13 -2A illustrates the transport with front
cover open. Fig. 13 -2B presents a detailed cutaway view of the transport
with a cassette in place and the tape threaded.
In order to perform the threading operation rapidly and accurately, the
capstan, audio shield, and female guide retract in unison when the tape
threading operation begins, and move into the normal operating position,
again in unison, when the tape is threaded. When the rewind and unthread
operation occurs, the same events take place with the additional step of
high -speed rewind. Rewinding is done without head -to -tape contact. The
design of the vacuum tape transport allows for gentle, rapid, and precision
tape handling in an enclosed, controlled environment, eliminating dust and
contamination from tape surfaces.
The vacuum columns, video and audio heads, and capstan are accessible
from the front for cleaning and maintenance. The carrousel load controls
are located on a panel between the two transports. The operator has access
to any group of four cassette bins in less than one second by pushing the
appropriate button on the load control panel. Cassettes are loaded from
the right-hand side of the carrousel through a double interlocked door
(Fig. 13 -3).
Control Console
The control console contains video, audio, signal, and waveform moni-
tors and four control panels: (1) automatic, (2) manual, (3) secondary,
and (4) maintenance. The control console also contains servo-card bays,
logic -card bays, the signal system, time -base -corrector electronics, and
power supplies. Many of the printed circuit boards used in the ACR -25,
such as signal- system and time -base -correction boards, are directly inter-
changeable with boards used in the AVR -1. In addition, the Mark XX video
head is used on both machines.
Automatic Control Panel
The automatic control panel consists of a left -hand and a right-hand
portion. The left -hand portion (Fig. 13 -4A) contains a 40 -event random -
access memory which can program the 24 cassettes in any number of se-
quences and events required up to 40. This portion also contains the se-
quence display, sequence buttons, event buttons. keyboard, and keyboard
display. The right -hand portion (Fig. 13 -4B) contains the warning panel
(similar to the AVR -1) and the status display, which tells the operator the
status of both tape transports and the carrousel. This portion also contains
a number of push buttons, as follows. The AUTO button establishes that
the machine will play cassettes in the sequence as determined by the pro-
gramming of the memory. The SEMI -AUTO button establishes that the
machine will play two previously threaded and cued cassettes in sequence
and will continue to play cassettes as they are manually selected by moving
e t.
11r111
ei
(A) Front cover open.
Fig. 13 -2. ACR -25

(B) Cutaway view.

cassette tape transport.
the bin selector switch to another number. The LOAD CONTROL button
allows access to the carrousel for the loading of cassettes. The READY
button threads and cues the first two cassettes to be played in an automatic
sequence. The CANCEL button returns any cassettes in the tape transports
to the carrousel. The AUTO PLAY button starts the playback of an automatic
sequence. The NEXT button immediately starts the tape next in line and
ends the tape that is playing. The REMOTE button delegates selected
functions to a remote point.

Fig. 13 -3. Loading of cassettes into carrousel.
Manual Control Panel

The manual control panel contains all controls required to load, thread,
cue, record, playback -shuttle, unthread, and unload cassettes from either
transport manually. This panel also contains an electronic tape timer and
push buttons for the various record modes, such as rewrite control track,
dub record, cue -sequence record, audio and video record ( transport A or
B or both) , etc.
(A) Left -hand section.
(B) Right -hand section.

Fig. 13 -4. Automatic control pond of ACR -25.
Secondary (Signal) Control Panel

The secondary control panel (Fig. 13 -5) is located below the automatic
control panel and contains the Mark IV Electronic Editor (optional acces-
sory) ; audio and video input and output level controls; standards selection
controls; tracking, chroma- level, and black -level controls; control -track re-
write (optional) ; and automatic playback adjust (optional) Except for
.
control -track rewrite and automatic playback adjust, all controls on this
panel exist on the AVR -1 and serve the same purpose.

Fig. 13 -5. Secondary (signal) control panel.
Maintenance (Setup) Panel

The controls on the maintenance panel (Fig. 13 -6) are those that are re-
quired to set up the machine initially for proper playback. They are identi-
cal to those controls found on the AVR -1, with the exception that there
are two of everything that requires adjustment (two transports, two heads,
etc.).
Spot Leader
The spot leader consists of a mechanical start -of -tape (sot) marker, 10
seconds of blank tape, one second of leader, up to 6 minutes of program
(at 15 in /s), one second of trailer, 10 seconds of blank tape, and the me-
chanical end -of -tape (eot) marker (Fig. 13 -7) . The sot and eot markers
are merely precision holes punched in the video tape.

Fig. 13 -6. Maintenance (setup) control panel.
Light Sensitive Markers
10 Sec 10 Sec
of Tape End of Tape I
Tape t
11
Program _ Blank Tape
Program Time
6 Minutes at 15 in/s l Maximum
12 Minutes at 7 -1/2 inls
10 Seconds Continuous Operation Minimum
Fig. 13 -7. Spot leader.
Fig. 13 -8 shows utilization of the cue track. The proposed SMPTE time
and address code (Chapter 10) has been incorporated in the ACR -25 and
is recorded on the cue track. The spare bits can be used to (A) verify pro-
gram content, (B) record automatic playback data (chroma phase, video
gain, audio gain, black level) , (C) provide a cue -up point for start of
program, (D) provide an auto -roll cue for a film projector or vtr, (E) pro-
vide an auto -roll cue for start of next cassette, or (F) provide a program
stop cue (Fig. 13 -9) .
Cassette
The cassette (Fig. 13 -10) contains a maximum of 6 minutes of program
material (at 15 in/s), plus leader and trailer. Programs may be recorded
onto a fully loaded cassette, or prerecorded tapes may be physically trans-
ferred from a reel to a cassette, thereby allowing complete interchange-
ability and avoiding further recording generations. The cassette spools
may be removed from the cassette and shipped or stored in a shipping
container.
The cassette is hinged and is easily opened for removal of the spools.
Captive record and rewind lock -out buttons are located on the cassette.
13 -2. AMPEX SYSTEM DESIGN CONSIDERATIONS

Development of the ACR -25, a broadcast -quality cassette video tape
recorder, required the design of many new electronic and mechanical
3lank Tape = Blank Tape
Start -of -Tape Cue Up Cue Track Auto Roll Other Au o Roll Next End -of -Tape
Marker Program Start BC Equipment Cassette Marker
Fig. 13 -8. Utilization of cue track.

systems. This section describes requirements and design solutions for the
signal -processing, headwheel- servo, and capstan -servo systems, including
their relation to conventional video tape recorders, as described by the
Ampex Corporation. In order to emphasize the reasons for particular sys-
tem parameters, a brief overview of the complete system is presented.
Overview
Most important in overall system design objectives were fully auto-
matic operation, convenient and reliable tape handling, and a cycle time
that would allow 10- second spots to be played back -to -back. To accomplish
these goals, the ACR -25 has two vacuum - threaded tape transports that are
supplied with video -tape cassettes from a shared, random -access carrousel
Leader 1 Second
-0.3 Sec 0.5 Sec 0.2 Sec-

Blank Tape Program
[1fID111111811 Illllllllllllllllllll
11111111111111111111111111111111111111 11111111111111111111111101111111111111V. umm11nnumnnnnnnnl
Control -Track Start Digital: Program-Content Verification, Cue -Up Point Program
Begin Play to Cue Color Setup Data Program Start Start
t- Trailer --
Program
Black
-Variaòle -0.2 Sec- Picture
A1111111111 1110111111M Iu111111111111. 11 111111111111111111111111. 11111111111
7
Control Cue Automatic Roll Cue Automatic Roll Cue Program
Track Track Start Other BC Equipment Start Next Cassette Stop
Fig. 13 -9. Control functions.
storage area. Transports, carrousel, high -power drive components, and pre-
amplifiers are contained in the transport console unit. Control logic, audio
and video processing, servos, and monitoring are contained in the control
console unit. A sequence of events determined by the control logic delivers
cassettes to the transports and threads the tapes in a reliable, high -speed
manner. Once threaded, the cassette recorder is much the same as any reel -
to -reel machine, so the burden then falls upon the standard vtr components
to provide high -quality, fully synchronous video. When each tape is no
longer needed, it is rewound, unthreaded, and finally returned to the
carrousel.
A much-simplified block diagram of the ACR -25 electronics is shown
in Fig. 13 -11. The basic approach to design of the dual- transport system
(A) Loaded cassette.
(B) Loading of tape.

Fig. 13 -10. Video -tape cassette.
Transport Servo
A A
I
Sh red
Control Fig. 13 -11. Simplified block diagram
Video Out Signal
Electronics
Logic of ACR -25.
Tran port Servo
was to share as much of the signal electronics as possible, but also to pro-
vide two complete sets of servo electronics, so that tape handling could be
accomplished as easily and rapidly as possible.
Signal System
Design considerations for the signal system included minimizing the
amount of circuitry required, self- contained operation with some optional
external circuit sharing, flexible record capabilities, and fast cycle -time-
Ch I
Video Head Preamp

A A
1 Ch 2
Signal System and Video

Time -Base Correction Out
Ch 7
Video Head Preamp
Ch 4
AlB Control L yi
Equal Control
DG Control i
C 1
o 0
DG Head B NJsrsas NNh
DG Head A
Fig. 13 -12. Video reproduce mode of ACR -25.

related responses combined with automatic correction features. These goals

were accomplished by adding a few simple switching circuits to a signal
system and time -base corrector that are electrically identical to those used
in the AVR -1.
In the reproduce mode, A/B switches are required for preamplifier rf,
differential -gain control, and equalization control, as shown in Fig. 13 -12.
Switching between transports is done in the uncombined head rf output
during a vertical interval. Since there are two video quadruplex heads, there
are eight video preamplifiers followed by a four -pole, double -throw elec-
tronic switch feeding the standard signal system. Switching is initiated at
a time determined by the control logic.
RF In From Amplifier
Head A
AIB Control-. Switch

Driver
Amplifier
and
Compensation
- RF Out to
Channel Amplifier
RF In From
Amplifier
Head B
Fig. 13 -13. Playback rf switch in ACR -25.
The four playback rf switches are constructed on two shielded printed -

wiring assemblies containing two switches each. A block diagram of one
switch is shown in Fig. 13 -13. Most critical of the specifications for this
switch is isolation, since it is possible for the video head in one transport
to be used for recording while the other is used for playback. During this
mode of operation, the signal present at the unwanted input is about 20 dB
higher than the nominal playback value. Rejection of the unwanted signal
is 80 dB, while the frequency response of the desired channel is -}0.1 dB
from 100 kHz to 16 MHz at unity gain. Second -harmonic distortion is
less than -60 dB.
Channel -by- channel equalization and differential -gain control are set to
particular values by dc voltages applied to the standard signal-system cir-
cuits. Since the two video heads have different characteristics, it is necessary
to switch these voltages obtained from two independent sets of front -panel
controls. Automatic control of equalization is an optional accessory as in
other vtr's.
Since there is only one demodulator, video signals off -tape are not avail-
able to one servo system during roll time if the other transport is being
ACR-25 ACR-25 or AVR -1

Uncorrected Video and
Operational Signals
i
Time-Base Corrected
'' I
Corrector Video Out
0-7 i
Electron ic Switch esi

J
Fig. 13-14. Sharing of time-base corrector.
used for playback. Therefore, the servos lock to the control track, and
when the A/B switch occurs, the new horizontal-sync time may differ by
as much as -±- 1/2 line. In order to produce a fully synchronous output, the
time-base corrector must have a 64-microsecond range. This wide time-
base-correction range is an extremely important factor in simplification of
the signal-system electronics for a dual-transport vtr. A further simplifica-
tion is possible since the time-base corrector is easily shared by two vtr's
as shown in Fig. 13-14.
Flexible record capabilities are provided by two switches in the record rf
path and one switch in the input video path, as shown in Fig. 13-15. In
record-only modes, the dub switch selects input video, while the two record
Signal System and

Video Head Preamp Playback Time-Base Correction
A Switch
Modulator Input
Video
Record Amplifie
A
Record RF
Record Playback Video

Levels ',Ann, Yvvs
Control
Video Head Pr eamp Logic
Record Amplifier
Record Switch
Record Levels
WV, VNIsAi
Fig. 13-15. Record mode of ACR-25.

switches direct the modulator output. It is possible to record on transport

A, transport B, or both. A block diagram of the dub switch is shown in
Fig. 13 -16. This switch isolates the two possible video sources without
degrading system specifications. Rejection of the unwanted signal is at
least 70 dB, while the frequency response of the desired signal is -?-0.1 dB
from dc to 10 MHz. Differential phase is within 0.5°, and differential gain
is less than 0.5 percent. Isolation specifications for the record switch are
determined by operation during the dub mode. Record rf leaking into the
active preamplifier must be avoided. This turns out to be relatively easy,
since the record amplifier is turned off and there is relay switching in the
preamplifier. With a gated limiter amplifier, the output from the undesired
External Video
From Input Filter - I
Resistive
mpedance
Matching
Dub Command H Switch

Driver
r Amplifier
and
Compensation
- Video Out to
Gain Control
and Modulator
Resistive
Time-Base
From Time -Base Impedance
Impedance
Corrector Matching
Fig. 13 -16. Dub switch in ACR -25.
channel is at least 70 dB below the incoming signal. Frequency response

through the desired channel is ±0.1 dB from 100 kHz to 16 MHz.
It is quite likely that spots to be played back -to -back will have been
made on different vtr's at different locations and with different variations
from standard practice. To handle these variations during automatic opera-
tion, chroma phase, video output level, and black level are controlled by
dc voltages which may be changed corresponding to digital information
recorded on the cue track.
Head and Capstan Servos
Each tape transport has associated with it a complete third -generation
head and capstan servo system. These two independent systems are tied to-
gether by control logic as required for back -to -back operation. Cycle time
(defined as the amount of time required after the completion of one spot
until the next spot is available) is almost entirely determined by the various
servos. Basic elements of cycle time are shown in Chart 13 -1. Although the
signal system needs only simple switches to handle back-to -back operation,
it is necessary for the head and capstan servos to respond to normal opera-
Chart 13 -1. Elements of ACR -25 Cycle Time

Thread Into Rewind Mode Thread
Stop Head and Tape Tape Into Top Columns
Move Capstan and Guide Move Capstan and Guide
Pull Tape Into One Column
Cue
Rewind Search
Unthread From One Column Park
Change Cassette Roll

Eject
Move Carrousel
Inject
tions more quickly and to perform some new functions. Other major com-
ponents of cycle time are determined by reel servos, the carrousel servo,
and various electromechanical components.
An important segment of cycle time is cuing. Each time a spot is to be
played, it is first threaded at a point ahead of the cue point. At the earliest
possible time, the tape is pulled forward at 150 in /s until digital informa-
tion from the cue track starts the park cycle. Fig. 13 -17 is a block diagram
of the cue servo. The basic functions of this servo are to generate a velocity
profile for parking and to determine when the tape is properly parked.
When the cue command is received, the up /down counter is preset to
488, which represents the number of capstan tachometer pulses between
Capstan
Weighted D -to -A
Converter Velocity
Reference
Parked
Capstan
Up -Down Counter
Tach Sicn
Park Preset
Up /Down
t
Capstan Capstan
Dir ction Motion
Cue
Fig. 13 -17. Cue servo in ACR -25.
where the park command will be given and the desired parking point.
Contained on the cue track is a continuous playback -speed countdown of
time before the start of message (som). In both 525- and 625 -line opera-
tion, the park command is given at two seconds, seven frames before the
som. Due to the difference in tape speed, a 525 -line tape will park at 0.2
second before som, whereas a 625 -line tape will park at 0.35 second before
som. These distances correspond to the respective roll times for 525- and
625 -line operation. Receipt of the park command releases the preset input
and allows the counter to function. Digital information from the counter
drives a weighted digital -to- analog converter which produces a reference
velocity profile for the capstan servo. Three conditions must be met before
the capstan is considered to be parked and the capstan servo is put into
the stop mode. These conditions are: the tape is within ± 1 frame of the
desired parking point, capstan motion is less than 3.75 in /s, and the cap-
stan direction is such that the last increment of motion will move it closer
to the desired parking point. The total time required for the cuing process
is less than one second when the tape is threaded at a point approximately
80 inches before som.
The time required for the headwheel to come up to speed and the time
required to stop are also important segments of cycle time. Start -up time
is mainly a function of motor -drive amplifier power and may be accom-
plished during the cuing cycle. Stopping is more important. Since the tape
is rewound under reel -servo control without any head -to-tape contact, it is
necessary to stop the headwheel prior to retracting the vacuum -guide block.
Circuitry was added to the headwheel servo to provide velocity control for
deceleration down to zero speed. With the brushless dc motor of the Mark
XX head, this was accomplished simply at a low level by merely reversing
the phase of the drive commands to the six -phase motor and detecting a
stopped condition in a manner similar to capstan parking.
Automatic features common to the AVR -1 are autotracking and guide
servo, which are helpful in back -to-back operation of tapes that vary from
standard practice. Even though the servos are capable of locking up on a
badly misplaced control track, this does not occur in the 0.2 second allowed
in cycle time. For this reason, one of the servo systems is equipped with
the capability of replacing any control track with one recorded to standard
specifications. This is done once when the tape is first previewed, and
thereafter fast lock-up will always be obtained.
Fig. 13 -18 is a block diagram of the control -track rewrite method.
Basically, the capstan position loop is locked by using an advanced control -
track read head so the control track can be erased and rewritten using the
normal heads.
Module B1 generates a positional -error signal by comparing the phase of
the advanced read -head signal to the headwheel-tachometer signal. Since
an incorrect control track may be of any phase, the head -tachometer pulses
may be shifted by 0 to 360° as determined by the control -track preset, a
front -panel adjustment. Both the advanced control -track head and the erase
head are located in the video -erase stack; hence a 5 -kHz low -pass filter is
required to eliminate erase -bias frequencies from the advanced control -
track signal. The control-track filter and squarer are standard circuits as
used in the normal playback mode. Position information from module B1
is added to the normal velocity information obtained by comparing the
headwheel tachometer and the capstan tachometer.
Advanced
/Control -Track
Head
Control -Track
Ver( Raf
Erase Head
Control -Track
Preset
Tach
-Mnn- 5 Hz
Low -Pass
Filter
Motor
0 -360°
Headwheel Control -Track Control-Track
Phase
Servo Record Filter
Shift
Headwheel
r
\
/Control-Track
Head
Ref
(Head Tacht
Ph se
Comp rator -- Squarer
Module B1
Co Capstan
Motor
MDA
+ Position Loop
Velocity
Loop
Tach
Frequency
Discriminator
Tape Motion
Fig. 13 -18. Control -track rewrite mode in ACR -25.
13 -3. THE RCA TCR -100
The RCA TCR -loo makes and /or plays tv program material contained
in cartridges which hold from 2 seconds to 3 minutes of tape. The machine
holds 22 cartridges (Fig. 13 -19) The system is most useful in handling
.
Control Panels
Play Transport Record Editor
no
11611§10 ®
nono
£3000 C3 oca
Setup Adjustments
Under This Door ll
Transfer _ Transfer
Station B
Station A
Tape Decks
Courtesy RCA
Fig. 13 -19. Front view of RCA TCR -100 cartridge system.
short program segments such as those that occur during a station break.
Commercials and program events in cartridge form can be aired auto-
matically in sequence, with all details, including threading and video
switching between cartridges in a sequence, handled by the system.
The cartridge system utilizes a video -tape cartridge as the carrier, or
package, for each recorded segment. It has the capability of playing back
the cartridges one after another, assembling signals from each into a com-
plete, continuous program. This is accomplished by including two playback
stations ( tape decks) within the equipment. Also included are appropriate
switching and control electronics and a cartridge -changer mechanism with
a capacity of 22 cartridges. The equipment is programmed to play a se-
quence simply by placing the appropriate cartridges in the changer mech-
anism in the order in which they are to be played. It is made ready by
activating a cue -up mode, and at the appropriate time, the sequence is
initiated by supplying a single start command.
The cartridge is a molded plastic container measuring approximately 21/2
X 31/3 X 5 inches. It holds two small spools of two -inch quadruplex video
tape. It provides proper tape handling in the equipment, as well as adequate
tape protection outside of the equipment. To achieve protection, doors on
three sides of the cartridge are normally closed. They open automatically
once the cartridge has reached the playing station of the equipment. For
convenience in handling and storing, there are no protrusions on the cart-
ridge; all requirements are met within the rectangular outline.
Start of Tape Start -of -Tape Marker End -of-Tape Marker End of Tape
236 ft Usable Tape Length

(Long -Play Cartridge)
Fig. 13 -20. Locations of mechanical cue markers.
Maximum tape capacity is 236 feet of usable tape length plus leader and
trailer. This gives a program playing time of a little more than 3 minutes
at 15 in /s for 525 -line television systems. The recorded format on the tape
is the same as for high -band quadruplex reel -to -reel tapes. There is only
one recording standard and one tape speed (15 in /s). The recorded format
also includes several cue marks, which are necessary for proper operation
of the system. These are placed on the tape automatically when a cartridge
is recorded.
Every cartridge has two permanent mechanical (reflective) cue marks
located near the ends of the tape (Fig. 13 -20) These are used to prevent .
the tape from being completely unwound under any condition while in the
equipment. Also, the start -of -tape marker is used as the bench mark for
recording. To make ready for recording, the equipment automatically
locates the start-of -tape marker in order to determine where on the tape a
recording should begin.
Start of Tape Start -of -Tape Marker Start -of- Program Cue End -of- Program Cue End -of -Tape Marker
Active
Leader Program t
1111 _ IIi01111IIIIfl11111111111111111 _: ó
5 Sec 2 Sec 2 Sec I 2 Sec
10 Sec
End of Tape
2 Sec Min 180 Sec Max
(Long -Play Cartridge)
Fig. 13 -21. Cartridge -tape format.

The recorded tape format ( Fig. 13 -21) requires two additional magnetic
tone marks on the cue track of the tape to determine the exact location
and duration of the actual program segment. A start -of- program tone mark
is placed exactly two seconds before the start of the active program material.
This mark is used to determine the exact point from which the equipment
will start when it is commanded to play back. Therefore, a two -second
preroll command must be given. The cartridge recorder will start up and
provide a fully synchronized color picture in less than two seconds so that
the signal may be reliably put on the air two seconds after start -up.
An eight- second end -of- program tone mark is placed ten seconds before
the end of the program. This mark is positioned during recording by telling
the system the duration of the program, from which the system computes
the correct location for the mark. There is also a mode available to reposi-
tion this mark to correct an improperly recorded tape. The trailing edge of
the eight- second tone mark is used for the start pulse to the next cartridge
in a sequence, so that the two-second preroll is automatically provided. This
gives the concept known as sequential automation.
The magazine is the cartridge- storage device and the physical interface
between the operator and the machine. The magazine contains 22 individual
bins for storing cartridges. The bins are mounted on a continuous belt
which can be indexed on command, clockwise or counterclockwise, to get
the desired cartridge to line up with the entry portals. Indexing is accom-
plished by a photoelectric sensor, which controls the actuation of a clutch
and gear -head motor.
To transfer the cartridge into the recorder, each entrance portal is
equipped with a transfer station containing pneumatically actuated claws
and a carriage. The claws engage the cartridge, and as the carriage travels,
the cartridge is pushed through the entry portal. To retract the cartridge,
the reverse procedure is followed, and the cartridge is pulled back into the
bin.
All of the bins are continuously accessible to the operator for loading
or unloading. However, the operator needs to know which bins are being
used. For this, small V- shaped, numbered flags are used. The position of a
flag indicates the status of the bin. If the flag is tipped to the right, it physi-
cally blocks the entrance to the bin, indicating that its cartridge has been
transferred into the recorder.
In addition to the mechanical portion, the magazine contains the logic
required to perform the basic functions of indexing clockwise or counter-
clockwise and transferring in or out. A photoelectric reader provides con-
tinuous information about the belt location to the recorder control system.
The two transports contain automatic threading mechanisms (Fig. 13 -22)
in addition to the normal transport elements. To provide long tape life, the
tape path and the threader have been designed to minimize contact with
the oxide side of the tape. In the tape path, the erase heads and guiding
elements contact only the back of the tape. A urethane -coated capstan is
used to provide the necessary tape drive force without the use of a pinch
roller; the capstan contacts only the back of the tape. Therefore, the only
elements that contact the tape on the oxide side are the video, control- track,
audio, and cue record /play heads.
Besides the usual signal electronics mounted close to the magnetic
transducers, the transport contains the control logic required to thread
and unthread the tape. Photoelectric sensors have also been included to
detect fail -safe reflective foil markers placed on the back of the tape near
the start and end.
Headwheel RIP Head
Guide Post
Back -Side Erase Audio Erase Head

Guide Post
Guide Roller
Capstan
Supply Reel Take -Up Reel
Cartridge
J
Fig. 13 -22. Tape threading path.
The video playback uses a standard quadruplex rotating head assembly.

In order to thread the tape through the vacuum guide and around the rest
of the tape path, a special mechanism extracts a loop of tape from the
cartridge and places the tape in the operating path. The threading mech-
anism then retracts so that it does not interfere with playback, record, or
rewinding of the tape in the normal tape path. At the end of a play or
record cycle, the tape is automatically rewound past the start -of- program
mark; then the threading mechanism extends and removes the tape from
the tape path, and the tape loop is drawn back into the cartridge. The
cartridge is then available to be removed from the playing station.
The cartridges are handled in the equipment by a changer mechanism
which can carry 22 cartridges (Fig. 13 -23). This mechanism consists of
a belt with 22 bins, and two transfer stations located above the entrance
to each playing station.
The transfer from the bin to the playing station is accomplished by a pair
of arms which rotate downward to engage the cartridge on its recessed
Transfer Station A Transfer Station B
Fig. 13 -23. TCR -100 belt changer mechanism.
plate at the rear. The arms then move horizontally back into the equipment
to place the cartridge into playing position. Upon completion of a play (or
record) cycle, the transfer arms again engage the cartridge and move it
back into the bin.
The control system in the equipment provides all the facilities for play-
ing cartridges in a continuous sequence. This sequence is determined by the
order in which the cartridges are loaded into the belt. Fig. 13 -23 shows the
numbering of the sequential locations (bins) in the belt. It is normal to
load the belt starting with bin 1. The belt is then positioned for starting by
Thread Cue A
Station A
Cued
/d I
Changer Belt
Advance -3 Advance +3 t
Play Cue toad 2
Thread Cue
Station B Cued
Fig. 13 -24. Operations sequence, play cue mode.

activating the home mode, which causes the belt to move so that bin 1 is
indexed at playing station A (left) .
The machine is made ready for playing a sequence by activating the play
cue mode. This initiates the sequence of operations shown in Fig. 13 -24,
which causes both the A and B playing stations to be loaded, threaded, and
cued. At the completion of the cycle, indicators tell the operator that the
play cue function is completed and the machine is ready to accept a play
command.
Play
Video Switch Video Switch
Station A -
r_Active
1---2 Sec
Program
SS
2
1
End Cue
Sec --i
Rewind
Unthread
Thread
Cue
Sf
Play
S
Active Program
EndI
3-i
1-10 Sec--I
Rewind
Unthread
Unload 3
Load
Thread
Cue 5
Changer Belt
Un oI'. 11
I
Advance
I Load 3
Advance -5
2 -1
External Cue
Advance 2
Advance +3
Unload 2 Load 4
Play
2 Sec
End Cue
- r-
Station B Cued
Thread
h-
I
Active Program 2 Cue
Fig. 13 -25. Operations sequence for three cartridges.
A more detailed description of the process is as follows. The play cue

command causes cartridge 1 to be loaded in station A ( the belt was already
positioned for this by the home mode). As soon as cartridge 1 is loaded, the
belt moves three spaces to the right, which positions bin 2 at station B.
The cartridge from bin 2 is then transferred into station B, and the belt
immediately shifts three spaces back to the left. This puts it in position to
receive cartridge 1 back in the bin after the cartridge has played and the
tape has been rewound and unthreaded.
As soon as each cartridge is in the playing station, the station auto-
matically threads the tape into the tape path and advances the tape in a for-
ward run direction until the start -of- program cue is located. The tape stops
accurately at this point, and the playing station is ready to accept a play
command.
Upon receipt of a play command, the A station is started in the play
mode (Fig. 13-25 ) The servo system provides a fully synchronized color
.
picture in less than two seconds, and the active program material com-
mences two seconds after start. The machine then plays through cartridge
1 until the end cue is detected ( starting ten seconds before the end of the
program). The trailing edge of the cartridge -1 end cue is directed to start
the B playing station. (This occurs two seconds before the end of the pro-
gram on cartridge 1.) Two seconds after the start of the B station, this
station will be fully synchronized, and the program is automatically switched
from the A station to the B station. The program switch takes place during
the vertical-blanking interval and gives a transition equivalent to an
excellent electronic splice.
As soon as the program switch to the B station has taken place, the A
station automatically rewinds cartridge 1 until the start cue is passed.
Cartridge 1 is then stopped, unthreaded from the tape path, and unloaded
back into bin 1 in the belt, which has been waiting in the proper position.
The belt then moves two spaces to the left, which brings bin 3 into position
at the A station. The cartridge from this bin is then transferred into the A
station, and the thread and cue cycle is again activated. When threading
and cuing of cartridge 3 are completed, station A is ready to accept a start
trigger. Meanwhile, as soon as cartridge 3 has been transferred, the belt
moves five spaces to the right, which places bin 2 at the B station, ready to
receive cartridge 2 when its cycle has been completed.
The B station meanwhile is continuing to play cartridge 2. When the
trailing edge of the end cue from cartridge 2 is passed, a start trigger is
directed back to the A station. This station, containing cartridge 3, begins
running, and after two seconds the program switches back to A. At this
point, the B station goes through the cycle of rewind, unthread, unload,
advance belt, load, thread, and cue to ready cartridge 4. The cycle of re-
moving one cartridge and readying the next one in the same playing
station takes a little less than twenty seconds. This means that the shortest
program segment which can be sequenced continuously is twenty seconds.
However, either the first or last segment in a sequence can be shorter, since
it is not necessary to complete the cartridge change cycle at the other
playing station during either the first or last segment of a sequence. In such
cases, the cuing system is capable of handling program segment lengths
down to two seconds.
Play Control
The play control panel is shown at the upper left in Fig. 13 -26. Preparing
the TCR machine for on- the -air operation is a 3 -step procedure: pro-
gramming, previewing, and playing. (1) To program the machine, the
number of sequences and the number of cartridges in each sequence are put
into the automatic AB sequence control, and the AUTOMATIC button is
pressed. The PLAY CUE -UP button readies the first two cartridges for play.
(2) To preview, the SEQUENCE PREVIEW button is pushed; then all events
in the upcoming sequence are seen. (3) To put the machine on the air, the
PLAY button is used. (If control is from another source, the REMOTE
CONTROL button is used.)
Courtesy RCA
Fig. 13 -26. Play and transport control panels.
Transport Control
The transport control panel is shown at the upper right in Fig. 13 -26.
These controls select the transport option in which the machine is to
operate and position the cartridge magazine. Provision is also made for
cartridge reject and resetting of logic functions. The left -hand portion of
the transport control panel is a 40 -item illuminated status indicator. Color -
keyed displays of red, orange, and green indicate relative importance. As a
result, the operator is always aware of what is happening inside the
machine.
E
MESSAGE END TIME SURI -NE SHIFT EXTEOURCE
PIT-POLL
RECORD CONTROL
Courtesy RCA
Fig. 13 -27. Record control panel.
Record Control
The record control panel is shown in Fig. 13 -27. Cartridge recording on
the TCR -100 is a three -step procedure: preset, cue -up, and recording. To
preset the machine, the message length is set on the MESSAGE END TIME
thumbwheels, and the external- source preroll time is set on the EXT SOURCE
PRE -ROLL register. Next, the RECORD CUE -UP button readies the cartridge
to be recorded. Then, a touch of the RECORD button starts the recording
process. Following completion of the recording, the machine automatically
rewinds and recues the cartridge for immediate replay. A touch of the
PLAY button rolls the tape for review.
System Electronics
I he system outlined in the preceding description could basically consist
of two independent automated vtr systems coupled together with a common
control system and a common cartridge changer belt. However, in such a
system, it is not necessary to provide two complete electronic systems for
the two vtr's because they are never used simultaneously. More specifically,
it is completely possible to time -share the video - signal electronics. This is
the most expensive portion of the tape-recorder electronics, so a major
economy can be realized.
The sharing of video -signal electronics in playback is accomplished by
performing the switching between the two playback stations while the
signal is still in the fm format as recorded on the tape (Fig. 13 -28A) .
Therefore, only one set of electronics- including demodulator, mono-
chrome and color time -base corrector, velocity and chroma error com-
pensator, dropout compensator, and signal -processing amplifier -is neces-
sary for the two playback stations. However, to allow for one playback
station to start while the other station is still on the air, it is necessary to
provide two complete servo systems. It is also necessary to have an fm play-
back preamplifier, an equalizer amplifier, and an fm switcher for each of
the two headwheel assemblies. With this equipment arrangement, the
system will provide instantaneous signal switching between the two tape
playback stations with excellent timing stability of both signals.
Since the video-playback signal electronics is identical to that in presently
available high -band video recorders, the ultimate economy is achieved
when the cartridge system "borrows" the signal electronics from an existing
reel -to -reel vtr (Fig. 13 -28B) . In this case, the cartridge system is a "slave"
to the reel -to-reel vtr, which becomes the "master." The master /slave ar-
rangement is accomplished by modifying the master machine to provide
an fm switch which selects between the normal fm output of the reel -to-
reel electronics and the output of the cartridge slave unit. This switch is
also arranged to operate during the vertical -blanking interval, and the servo
system of the master machine is modified so that an excellent video switch
transition is possible between the master machine and a cartridge. This
r r
He dwheet Headwheel
A sembly Assembly
_1_1 T T
4-Cha nel 4-Cha nel
Se vo
lifers Servo
Am Amplifiers
Sys em System
and witcher and Switcher
Mono & Color Video

Demod Timing Proc
Correction Amp
Dropout Velocity and

Comp Chroma Comp
(A) Cartridge playback.
Headwheel Headwheel
Assembly Assembly
4-Channel 4-Cha nel

Se vo Se vo
Am liliers Am lifers
Sys em System
and Switcher and Swi cher
Slave Unit A B
Slave
Mono & Color Video
Headwheel Demod Timing Proc
Assembly Correction Amp
Master
Dropout
Comp
4-Channel
Se vo
Am Idlers
System
and Switch r
(B) Slave/master machine electronics.

Fig. 13-28. System arrangements for TCR-100.
means that it is possible to insert a sequence of cartridges in a program

played on the master reel -to -reel machine with good video switch transi-
tions at all points. It is also possible to go from the master machine to the
cartridge playback, change or recue the tape on the master machine while
cartridges are playing, and then go back to the master machine at the end
of the cartridge sequence, all within the one master /slave equipment
complex.
The TCR -100 electronics systems occupy enclosed spaces above and
below the cartridge magazine and transport section. All of the control
electronics is in a roll -out drawer behind the control panels above the
magazine. Fig. 13 -29 illustrates the roll -out and swing -down electronics
Courtesy RCA
Fig. 13 -29. Control electronics in servicing position.
arrangement for servicing accessibility. The electronics is made up of

printed- circuit cards which plug into nests in each electronics drawer.
Signal -handling and transport servo- system electronics is in a drawer below
the magazine.
Normal operation of the system requires that a setup routine be per-
formed periodically, probably daily in most installations. In this routine
procedure, the playback system of the cartridge recorder is equalized for
optimum playback of a standard test cartridge. With the playback system
thus adjusted, the recording function of the machine is then checked and
adjusted, if necessary, to insure that the playback system and its automatic
Courtesy RCA
Fig. 13 -30. Setup adjustneant panel with door open.
Courtesy RCA
Fig. 13 -31. Rear view of TCR -100.
correction devices are still in center range. Fig. 13 -30 illustrates the setup
control panel ( the narrow panel above the cartridge magazine) .
Both tape decks roll out on slides to allow routine maintenance pro-
cedures. Access to the transport decks is also available from the rear. Fig.
13 -31 is a rear view of the TCR -100.
EXERCISES
Q13 -1. Can tape recorded on a cassette/cartridge system be played back on a

conventional vtr?
Q 13 -2. Can tapes recorded on a conventional vtr be played on the cassette/
cartridge machine?
Q13-3. Does the Ampex ACR -25 cassette machine work alone, or in con-
junction with another machine?
Q13 -4. Does the RCA TCR -100 cartridge machine work alone, or in con-
junction with another machine?
Q13 -5. What is the meaning of (A) sot and (B) eot?
Q13 -6. What means is used to mark sot and eot on (A) Ampex cassette tape
and (B) RCA cartridge tape?
Q13 -7. If the cassette/cartridge machine is self- contained, give the number of
(A) video head panels, (B) transports, (C) demodulators, and (D)
servo systems necessary.
Q13 -8. What must be the maximum allowable lock -up time for a cassette/
cartridge tape system?
CHAPTER 14
Video Disc Recording Systems
The video disc recorder, which essentially uses the same basic recording
and playback techniques as the magnetic tape system, is rapidly becoming
standard equipment in the television broadcast industry. The initial appli-
cation of this system was in the slow-motion, stop- action type of "instant
replay," and this is still one of the major functions performed by video
disc recorders. However, its use has expanded in recent years into the field
of providing automatic and practically unlimited varieties of special effects,
editing, and animation techniques for creative television. Its possibilities
in special teleproduction applications are rapidly expanding and are limited
only by the imagination of directors, producers, and artists.
We will first examine the basics of the single -disc, two -head system. Then
we will progress to the Ampex two -disc, four-head system, which is the
most commonly used video disc recorder at the time of this writing. If the
reader will study the contents of this chapter thoroughly, he should be able
to keep abreast of future developments in this field, since certain funda-
mentals serve as the foundation for all such systems.
14 -1. THE SINGLE -DISC, TWO -HEAD VIDEO SYSTEM

Examples of the single -disc, two -head system are the Visual Electronics
Corp. VM -90 and the VDR -1000 by Data Memory, Inc., and TeleMation,
Inc. Fig. 14 -1 illustrates the basic function of a two -head disc recording
system. The highlights of this technique can be outlined as follows:
1. The disc consists of aluminum substrates lapped to optical flatness,

then electroplated with a thin layer of magnetic nickel -cobalt. This
surface is then plated, or flashed, with a few microinches of rhodium
to provide a mirror -like running surface for the heads and to prevent
surface corrosion. The DMI video disc has a 1/4 -inch aluminum base
coated with polished copper and an overlay of nickel -cobalt alloy
482
VIDEO DISC RECORDING SYSTEMS 483
Fig. 14 -1. Block diagram of two -head video disc recorder.

which accepts the magnetic signals. A coating of Chromecote pro-

vides an extremely smooth riding surface for the heads.
2. The rotating disc is contacted by two record -playback heads, one on
top and one on the underside.
3. Individual fields are recorded as separate, circular tracks on the disc
surface. In Fig. 14 -1, the odd -numbered fields are recorded on the
top, and the even-numbered fields are recorded on the bottom
4. The recording sequence is as follows: Starting at the inside edge of
the disc, field 1 is recorded by the top head. Field 2 is recorded by
the bottom head. During the time of field 2, the top stepper motor
advances the top head two tracks. The top head then records field 3.
During the time of field 3, the bottom motor advances the bottom
head two tracks. The entire process is then repeated.
5. At the end of travel in the first direction (inside -out) , the last two
fields are recorded, and each head moves back one step, again record-
ing the odd field on top and the even field on the bottom. The
steppers then revert to movement to every other track, and informa-
tion is recorded on the tracks left vacant on the first pass. With this
two -head system, all tracks are filled in 30 seconds.
6. For slow- motion or stop- action playback, a field must be used over
and over. To produce interlaced frames, the video information in a
single field is first delayed by a half line and then (by means of an
electronic switch) is undelayed for the next field duration. The in-
formation in the two fields is identical, but displaced vertically one-
half line on the raster to provide a full interlaced frame. The question
usually asked here is why a full frame is not repeated rather than
one field. The answer is that in scanning the two fields making up a
frame, a time differential of 1/60 second exists at any one point on
the raster. A rapidly moving object (such as a football, for example)
will produce a very noticeable double-exposure effect in the interlaced
frame.
14 -2. THE AMPEX HS-100 VIDEO RECORDER /REPRODUCER
The Ampex HS -100 is a two -disc, four -head recorder /reproducer system;
it is quite popular for a large number of special applications at the time of
this writing. We will first explore the general features of the equipment,
and then progress into more detailed analysis of circuit functions. Much of
the electronics is the same as that described for the conventional quadruplex
recorder. We will, therefore, emphasize the circuits that differ from those
already covered.
General
The Ampex HS -100 slow-motion video recorder and reproducer (Fig.
14 -2) is a transportable instant -replay television recorder /reproducer man-
Disc & Head Assembly
Disc Servo Unit

IDSUI
Electronics Signal Unit

IES UI
Output Processing Unit

IOPUI
Control Unit

Fig. 14 -2. Ampex HS -100 video disc recorcer¡ reproducer system.
ufactured for use in studios, mobile vans, or indoor remote broadcast sites.
It is capable of recording standard NTSC color or monochrome video
signals and then immediately replaying the recorded material, either for-
ward or in reverse, at normal speed, twice normal speed, one-half normal
speed, one -fifth normal speed, or at manually controlled, continuously
variable speeds ranging from stop- action (freeze) to normal. The HS -100
also permits operator-controlled, single -frame video advance, either forward
or in reverse, whenever playback is in the freeze mode. The video output
fully complies with NTSC and FCC standards regardless of speed or direc-
tion changes. Storage capacity is 1800 television fields, corresponding to 30
seconds of video material in the normal record mode, or 60 seconds of
material in the alternate-field record mode ( i.e., recording only every other
television field).
The HS -100 is packaged in four units: the disc servo unit ( dsu) , the
electronics signal unit (esu) , the output processing unit (opu) , and the
control unit. Each unit is enclosed in a weather -resistant metal cabinet.
Carrying handles on the cabinets facilitate movement of the equipment
during transit and installation. For operation, the three larger units (dsu,
esu, and opu) can be stacked vertically with the dsu on top, the esu in the
center, and the opu on the bottom; or they can be placed side by side with
the esu in the center. All controls required to operate the HS -100 are on
the control unit. The control unit connects to the esu through a single 30-
foot multiconductor cable (200 -foot cable extensions are available on
special order) and can be placed in any convenient location. Features of
the individual units are described in the following paragraphs.
Disc Servo Unit (DSU)

The disc servo unit contains the electromechanical components of the
HS -100 and their associated electronics. Electromechanical components in-
clude a disc drive assembly, with its associated circuitry, and four stepper
assemblies. The disc drive assembly controls the rotation of two magnetic
recording discs, each of which provides two recording surfaces (top and
bottom side of each disc). Each of the four stepper assemblies controls
the movement of one carriage assembly, with its record /reproduce head,
across one of the four disc surfaces. The disc drive assembly and the four
stepper assemblies are mounted on a machined aluminum top plate, which
is in turn shock -mounted to the frame of the dsu cabinet. The top plate
and the electromechanical and electronic parts mounted on it are called,
collectively, the top -plate assembly. Protection for the top -plate assembly is
provided by a clear acrylic dust cover. During handling and shipping,
additional protection is provided by a removable steel cover that fits over
the acrylic dust cover and latches to the dsu cabinet.
The electronics for the dsu is on plug -in printed circuit modules (cards)
which plug into a card rack extending across the front of the unit chassis
below the top -plate assembly. These components are protected by a door
on the front of the unit. A connector panel and an air exhaust opening are
located at the rear of the unit.
Electronics Signal Unit (ESU)

The electronics signal unit contains most of the signal- system electronics,
the control logic, and the major parts of the power- supply system. Printed
circuit modules on plug -in card assemblies are housed in a card rack ex-
tending across the front of the unit. A panel above the card rack contains
secondary operating controls. A power supply, containing front -mounted
fuses and test points, is mounted on a second panel directly below the card
rack. The front of the esu is protected by a hinged metal door. A connector
panel at the rear of the unit contains connectors for the main power cable,
video inputs and outputs, sync inputs, and cables that interconnect with
the disc servo unit, the output processing unit, and the control unit.
Cooling for the esu circuits is provided by an internal blower.
Output Processing Unit (OPU)
The output processing unit contains a standard Ampex Amtec, a stan-
dard Ampex Colortec, and an Ampex processing amplifier. These assem-
blies are mounted on slide rails in the opu cabinet for easy maintenance
access. All normally used operating controls (used for initial equipment
setup) are on the front of the individual assemblies. The cabinet is equipped
with hinged front and rear doors, a connector panel, and a blower. Electri-
cal connections to the opu are made at the connector panel at the bottom
rear of the unit.
Amtec Time -Element Compensator -The Amtec time-element (time -
error) compensator is a self- contained electronic assembly. Most of the
Amtec circuits are contained on plug -in printed wiring board assemblies
which can be easily removed when the Amtec assembly is pulled forward
on its slide -rail mountings. All operating controls for the Amtec are
mounted on the right side of the assembly front panel.
Colortec Direct Color Recovery System -The Colortec direct color re-
covery system is, like the Amtec, a functionally self -contained electronic
assembly, utilizing plug -in printed wiring boards. All operating controls
for the Colortec are mounted on the right side of its front panel.
Processing Amplifier -The processing amplifier is physically similar to
the Amtec and the Colortec in that it employs solid -state circuitry on
printed wiring board assemblies. Operating controls and test points are
located on the right side of the front panel.
Control Unit
The control unit contains all the primary controls used to operate the
HS-100. It connects to the electronics signal unit by means of a multi -
conductor cable. All controls are illuminated push buttons except for a lever
used for variable slow- motion speed control. All internal circuitry is
mounted on one printed wiring board assembly. Mounted above the con-
trol panel is an illuminated clock -type dial, calibrated from 0 to 30. A
white pointer on this dial indicates the head location relative to the 30-
second storage capacity of the system. A red pointer tracks with the white
pointer. This pointer can be stopped and then reset to its tracking position
at the option of the operator to provide a cue marking. A red on -air lamp
is provided at the top of the control unit. This lamp can be wired into the
studio master control panel if desired.
14 -3. BASIC FUNCTIONAL DESCRIPTION OF HS -100
Video recording in the HS -100 is accomplished by using high -band

standards. The incoming composite video signal is converted to a fre-
quency -modulated signal before it is recorded. During playback, the signal

is demodulated to produce a standard video output signal. The record and
reproduce processes are described separately in the following paragraphs.
Record Process
In the HS -100, video signals are recorded on the four surfaces of two
metal discs rotating about a common vertical shaft. Recording is con-
tinuous until the operator overrides the record mode by selecting a repro-
duce or fast search mode. As long as recording continues, the latest 30
seconds (60 seconds for alternate-field recording) of recorded video is
maintained in storage, ready for instant playback. Material recorded prior
to the 30- second storage limit is progressively erased to permit recording
of new material.
Disc Rotation-The rotation speed of the two recording discs is 60
revolutions per second (3600 rpm) This speed is precisely controlled by
.
a disc -drive servo system which instantly detects and corrects drive -motor
speed variations (Fig. 14 -3) . The primary purpose of the disc -drive servo
is to lock the rotation of the discs in phase with the external reference
vertical sync. This phase lock ensures that each complete revolution of the
discs corresponds exactly to one television field, beginning and ending
during the vertical blanking period. The disc -drive servo comprises an
optical tachometer, driven by encoders on the disc drive shaft, the electronic
circuits associated with the tachometer, a velocity discriminator, a phase
comparator, the motor drive amplifiers, and the disc drive motor.
Head Transducer
To/From Head (1of 41
Preamplifier
r
Carriage
End Stop Stepping
Sensors Motor
(1 of 41
Car jage Tach and

Disc Drive
Drive Control Index Encoders Shaft
and Amplifier Processor
i
Logic and Disc
Composite
Swit hing Drive
Sync Input
Circuits Motor
1
Disc Motor
i Drive Control
andAmplifiers
Operator Co trol (Phase and
Commands Unit Velocity)
Fig. 14 -3. Simplified diagram of disc -drive and carriage- stepping systems.
Head Movement -Each of the four head assemblies used in the HS -100
is moved by an independent stepper assembly that steps the head radially
across the surface of the disc. For purposes of identification, the heads and
their associated signal paths are referred to as head (or channel) A, head
(or channel) B, head (or channel) C, and head (or channel) D. In addition
to recording, each head also functions as a playback and erase head. During
operation, head A steps radially across the top of the top disc, head B
steps across the bottom of the top disc, head C steps across the top of the
bottom disc, and head D steps across the bottom of the bottom disc.
Each head assembly consists of a ferrite head transducer and two ferrite
pads mounted so that the head and the two pads extend perpendicularly
from the corners of a triangular platform. The head and the two pads pro-
vide a stable three -point contact with the disc surface, and they are held
against the disc by a cantilever spring which bears against the rear of the
platform. The head carriage assembly, which moves the head assembly, is
mounted on guide rails extending radially across the disc surface. As the
head carriage is moved a given distance along the rails, the head transducer
is carried across the disc in the same direction and for the same distance.
Driving power for the carriage assembly is provided by a stepping motor
controlled by logic circuits and by end-stop sensors ( which provide revers-
ing commands) The carriage assembly is coupled to the shaft of the
.
stepping motor through a pinned stainless -steel belt. The manner in which
the stepping motors move (or step) their associated carriage assemblies is
determined by the track format for the mode of operation being used.
Track Format, Normal Record Mode -Each field is recorded on the disc
as a circular track; the head is held stationary while the disc makes one
complete revolution. When head A completes recording a single field,
head B starts recording the next field. While head B is recording, head A is
being stepped to a new position. When head B has recorded its field, head
C records the next field. While head C is recording, heads A and B are
both being stepped to new positions. When head C has completed record-
ing one field, head D starts recording, head A erases the track in which it
is now positioned, and heads B and C are stepped to new positions. When
head D completes recording one field, head A starts recording in the track
it just erased, head B erases, and heads C and D step to new track locations.
In this manner each head records every fourth field, and successive fields
are recorded by heads A, B, C, and D in rotational sequence. Heads A and C
record odd -numbered fields; heads B and D record even -numbered fields.
Recording, moving, and erasing follow a definite sequence for each
head. For example, during field 1, head A records. During field 2, the
stepper assembly moves head A 0.010 inch radially across the disc ( tracks
are 0.007 inch wide, 0.010 inch center -to- center) During field 3, head A
.
is moved an additional 0.010 inch radially, placing it a distance of two

tracks away from where it recorded (during field 1). During field 4, head
A erases, with a dc current, any signal previously recorded on the track in
TELEVISION BROADCASTING
Table 14 -1. Head Sequencing in Normal Record Mode
Head Field Field Field Field

(or 1, 5, 9, 2, 6, 10, 3, 7, 11, 4, 8, 12,
Channel) 13, etc. 14, etc. 15, etc. 16, etc.
A Record Step Step Erase

B Erase Record Step Step
C Step Erase Record Step
D Step Step Erase Record
which it is now positioned. At the start of field 5, the erase current to head
A is switched off, and head A is fed the fm signal output of the record
amplifier ( part of record signal electronics) The sequence for each head .
is shown in Table 14 -1.

Carriage Reversing- During any given field, one head is recording, one
head is erasing, and the remaining two heads are being stepped to new
track positions. The heads move in this manner toward the center of the
discs until head A eventually reaches its innermost track. This position of
head A is sensed by a lamp and photocell arrangement positioned so that
the photocell detects head A as it makes its first step from its last record
track position. The photocell output, acting through the carriage -control
logic circuits, prevents head A from making the second stepping movement.
During the two subsequent fields, heads B and C also reach their innermost
point of travel and are similarly prevented from making their second
stepping motion. On the next field, head A records on the track where it
was stopped, and head D steps one track. On the following field, the
direction of rotation of all four stepping motors is reversed. The heads
then begin stepping toward the outer edge of the disc, following the normal
sequence shown in Table 14 -1 (Table 14 -2 for the alternate -field mode)
and recording between the tracks used when the head movement was
toward the center of the discs.
At the outer edge of the discs, head travel is again inhibited and reversed
by a second lamp and photocell arrangement. Thus, the heads travel con-
tinuously until stopped by an operator command. If a stepping error occurs,
Table 14 -2. Head Sequencing in Alternate -Field Record Mode

Field Field Field Field Field Field Field Field
Head 1, 9, 2, 10, 3, 11, 4, 12, 5, 13, 6, 14, 7, 15, 8, 16,
(or 17, 25, 18, 26 19, 27, 20, 28, 21, 29, 22, 30, 23, 31, 24, 32,
Channel) etc. etc. etc. etc. etc. etc. etc. etc.
A Record Step Step Erase Erase

B Erase Erase Record Step Step
C Step Erase Erase Record Step
D Step Step Erase Erase Record
Blanks indica e periods of inattivi y.
the head-carriage logic circuits detect it and correct it at the end stops
before allowing the carriages to reverse direction.
Alternate -Field Recording -In the alternate-field recording mode of
operation, head sequencing is the same as for the normal recording mode,
except that it is triggered at half the rate. The rotation speed of the discs,
the track width, and the track spacing are unchanged. The sequence is
shown in Table 14 -2. Note that only the odd -numbered fields are recorded.
During even -numbered fields, no steps are triggered, and the record
amplifier is turned off. Erase current, however, is permitted to flow during
the even-numbered field intervals.
Record Signal Path -The record signal path, shown in Fig. 14 -4, con-
tains a modulator (located in the electronics unit) and a record amplifier
and four preamplifiers (located in the disc servo unit) The modulator
.
module contains a preamplifier circuit which amplifies the composite video

input, a pre-emphasis network, and a modulator which converts the com-
posite video to a wideband fm signal (rf) .
In the record amplifier module, the wideband fm is amplified and divided
into four sequentially gated outputs, each of which is applied to a different
one of the four preamplifiers. Each preamplifier contains a relay which is
energized during the record mode and de- energized during playback.
During recording, signals gated out of the record amplifier are applied
directly to the head transducer by way of a record -current level -adjustment
potentiometer (in the preamplifier) and the preamplifier relay contacts.
An erase circuit in each preamplifier supplies a dc erase signal to its associ-
ated head transducer when commanded to do so by a gating pulse. The
same pulses that gate the fm signals out of the record amplifier supply the
erase commands for the preamplifiers. The lines carrying the gating pulses
are connected in such a way that the pulse that gates the channel -A signal
out of the record amplifier turns on the erase function in the channel -B
preamplifier. Similarly, the pulse that gates the channel -B signal out of the
record amplifier simultaneously turns on the erase function in the channel -C
preamplifier. This sequence, applied to all four channels, ensures that when
one head is recording, the head that will record next is erasing previously
recorded material.
Reproduce Process
Among the factors involved in the playback of the recorded material
are speed, direction, line interlace, chroma phase, and alternate -field play-
back. These factors are discussed in the following paragraphs.
Normal Speed -In the normal -speed, forward -direction playback mode,
the sequence of carriage movement is identical to that used in record. The
head connections are transferred, by means of relays, from the record and
erase amplifiers to the reproduce preamplifiers. The outputs of the repro-
duce preamplifiers are sequentially gated through the rest of the reproduce
electronics in the same manner that the record current was gated to the
11
E S°aÉ
o
9_
oi g1Sd
5,,
r.
E
á
óg .ti"
-
1
w á 2
.
i12%
p
,=L
0 3 w W I4o
Fig. 14 -4. Simplified block diagram of record /reproduce signal path.

heads in the record mode. The sequence of operations is shown in Table

14 -3. Each field is reproduced in the exact sequence in which it was re-
corded so that the demodulated video output is in standard NTSC form.
-In
Still -Frame the still- frame, or freeze, mode, the playback sequence
of Table 14 -3 is stopped on a particular field, and the system video output
is derived from the continuously repeated playback of a single track. In
this mode, line interlace and chroma phase are restored by special tech-
niques to produce a standard television signal.
Line Interlace -The normal television video signal is a succession of
odd and even fields characterized by a half -line shift of horizontal sync
(with respect to vertical) in each field (Fig. 14 -5) This half-line shift
.
produces line interlace of the two fields that constitute a frame. When, as
in the case of still framing, each successive field is derived from the same
recorded track, and is therefore identical to the one preceding it, inter-
lace must be restored artificially.
The phasing of the record switcher in the HS-100 is such that each
recorded field begins and ends just after the last equalizing pulse of the
-
vertical interval (point A' or B' of Fig. 14 -5) Odd fields, as recorded on
.
A A'
L.._
Fig. 14 5. Record switcher phasing.
the disc by heads A and C, begin at A' and end at B'; even fields, as
recorded by heads B and D, begin at B' and end at A'. To produce line
interlace artificially, odd fields are changed to even fields, or even fields are
changed to odd fields, by insertion of a half-line delay in the video signal
during the horizontal- scanning interval of each field (i.e., from A' to B,
or from B' to A) . The half-line delay insertion is controlled by the system
logic. By knowing what type of field is required (by examining reference
sync) and knowing what type of field is being reproduced by each head,
the logic controls the insertion of the half-line delay as required, always
removing the half -line delay during the vertical interval, A to A' and B
to B'. Accordingly, during still- framing, the half -line delay is inserted
during the horizontal- scanning interval of alternate fields.
Chroma Phase -In the NTSC color system, the frequency relationship
between the chroma subcarrier and the horizontal and vertical scanning
rates is such that the chroma phase advances 180° each line and each frame
(dot interlace) This minimizes chroma-luminance cross talk, since the
.
effects are reversed on successive scans.

In still- framing, a chroma -phase problem arises from attempts to gen-
erate a continuous signal from a single recorded field. In scanning a com-
plete field, the chroma at the end of the field is advanced 90° with respect
to its phase at the beginning of the field. If the field is then rescanned from
the beginning, a 90° phase discontinuity appears in the chroma signal at
the beginning of the scan. This would not only destroy dot interlace, but
in a normal receiver would seriously disrupt the color demodulation process.
The chroma phase shift is further influenced by the insertion or removal
of the half-line delay. Insertion of the half -line delay retards the chroma
phase 90 °; removal of the delay advances the chroma phase 90 °. Thus,
when the half-line delay is inserted at the beginning of a rescan, its 90°
phase shift adds to the 90° shift caused by rescanning, producing a total
chroma phase shift of 180 °. Conversely, if at the beginning of a rescan
the half -line delay is switched out, its phase shift cancels out the 90° shift
caused by rescanning. The combined result when still- framing, therefore,
is that a 180° shift occurs in the chroma phase at the beginning of every
second field. This effect is compensated for by a chroma inverter which
extracts the chroma signal ( including burst) from the composite video
playback signal, reverses its phase each time the half -line delay is inserted,
and recombines it with the luminance portion of the signal.
Slow Motion-Slow motion is essentially a combination of normal
motion and still- framing. To produce the effect of slow motion, each
recorded track is scanned not once but several times, depending on the
slow- motion rate selected, after which the playback signal is taken from
the next track. Selection of a particular speed determines the average num-
ber of scans per track, even though some tracks may be scanned more often
than others. For example, if a speed reduction of 2 to 1 is selected, each
track is scanned twice; at a 3 -to -1 reduction, each track is scanned three
times. At a 2.5 -to -1 speed reduction, half the tracks are scanned twice and
half are scanned three times. Thus speed control is continuously variable
from normal to freeze.
During the time a particular track is being rescanned, the system oper-
ates exactly as described for still- framing (freeze mode) . Carriage motion
stops, the signal is derived from one particular head, the half-line delay is
switched in or out at the beginning of each rescan, and the chroma phase
is reversed each time the half -line delay is switched in. When the playback
signal is advanced from one track to the next, carriage movement and
head switching progress from one field to the next as in normal motion
(Table 14 -3). Since switching from one track to the next produces a
normal transition from one field to the next, the state of the half -line delay
and chroma inverter remains unchanged during the transition. That is, if
the half -line delay was in the signal path before the switch, it remains in
VIDEO DISC RECORDING SYSTEMS i95
Table 14 -3. Head Sequencing in Normal Reproduce Mode

(or 1, 5, 9, 2, 6, 10, 3, 7, 11, 4, 8, 12,
Channel) etc. etc. etc. etc.
A Reproduce Step Step

B Reproduce Step Step
C Step Reproduce Step
D Step Step Reproduce
after; if it was bypassed before the switch, it remains bypassed after. Simi-
larly, no inversion of the chroma phase is carried out at this time.
Reverse Motion In
reverse -motion playback (Table 14 -4), the se-
quence of carriage motion and head switching is reversed from that shown
in Table 14 -3, and the carriages are made to move in the opposite direction.
Thus the fields are played back in a sequence opposite to that in which they
were recorded.
The head switching sequence, D-C -B -A, preserves the normal progression
of fields from odd to even, but loses the track -to -track phase continuity of
the chroma signal. In switching, for example, from head D to head C,
switching is from the end of one field to the beginning of the one which
preceded it in the original recording. This constitutes a 180° chroma -phase
reversal, which must be corrected by reversing the chroma phase in the
chroma inverter. Thus, in switching from track to track in the reverse -
motion direction, the half -line delay is not altered, but the chroma phase
is reversed by the chroma inverter.
Table 14 -4. Reverse -Playback Head Sequencing

(or 1, 5, 2, 6, 3, 7, 4, 8,
Channel) etc. etc. etc. etc.
A Reproduce Step Step

B Step Step Reproduce
C Step Reproduce Step
D Reproduce Step Step
For rescanning a track in slow-motion reverse, the action of the half -line
delay and chroma inverter is identical to that for rescanning in the forward
direction.
Alternate -Field Playback-Recordings made in the alternate-field mode
differ from those made in the normal mode in that only odd fields are
recorded ( the first field of each frame) . Since all fields are odd, it is neces-
sary to change the state of the half -line delay (i.e., either insert it or remove
it, as the case may be) each time the signal is switched from one track to
the next, both in the forward and reverse motion directions. In going
These signals are fed to the processing amplifier, which combines the
composite video and the time -corrected burst (added to back porch of
composite video) to provide a time -corrected composite video signal. The
processing amplifier provides two composite video outputs, one non-
composite video output, and one sync output.
Fast Search
Fast search is used to move the heads rapidly (at about four times normal
speed) from one point on the discs to another. In fast search, as in normal
operation, the heads must remain precisely in step; otherwise loss of field -
to -field continuity would result in subsequent playback. Therefore, the
sequence of motion is kept the same as in normal -speed operation. Be-
cause of the inertia of the carriage drive system moving at search speed,
it is not convenient to reverse the direction of travel of the carriages at
the inner and outer limits of travel. Therefore, a lamp and photocell ar-
rangement, located on carriage drive A, detects the approach of the heads
to the inner and outer limits and briefly slows the carriage speed to normal
while the carriage direction is being reversed.
14 -4. THE HS -100 VIDEO SIGNAL SYSTEM

The functional block diagram of Fig. 14 -6 represents the signal system
in somewhat more detail than was shown in Fig. 14 -4. The signals recorded
on the HS -100 are composite video signals and may be either NTSC color
or monochrome. The video system has two modulation systems, fm for
recording and playback, and a -m for the half-line delay function. The fre-
quencies of the fm modulator are 7.06 MHz for tip of sync, 7.90 MHz for
blanking, and 10 MHz for peak white. The video system consists of four
head preamplifier assemblies, each located on a carriage drive assembly, an
rf switching logic module and a record amplifier module located in the
dsu, and eight signal -processing modules located in the esu.
Modulator Module
The modulator module contains circuitry to provide feedback -stabilized
amplification and dc restoration of the composite video signal before it is
applied to the fm modulator. A fast -switching voltage -controlled multivi-
brator is used to produce frequency modulation with low distortion. The
frequency-modulation process is similar to that used in video -tape recorders
in which the video information is contained in the form of sideband
energy. The carrier frequency is the same as that used in the high -band,
525 -line Ampex VR -1200 and VR -2000 video -tape recorders.
Record Amplifier Module

The record amplifier functions on the same principle as that of conven-
tional quadruplex systems except that fm is gated and switched to the heads
during the record mode rather than being fed to all four heads simultane-
ously. One head is always recording, one head erasing, and two heads mov-
ing to new positions.
The record amplifier receives a 0.7 -volt (peak -to -peak) fm signal from
the modulator. The signal is amplified to 2 volts peak to peak, which is a
level suitable to pass through switching circuits and provide sufficient
power to drive the recording heads to saturation of the recording medium.
The signal is separated into four output channels under control of four
gating signals.
RF Switcher Module
The rf switcher and switching logic module in J12 of the dsu receives
four rf inputs, one from each of the head preamplifiers, and five control
logic signals. The four preamplifier outputs are gated through the rf
switcher for recombination, either sequentially or on an individual-field
basis, depending on the selected reproduce mode. The module also contains
the circuits generating the gate signals for record, erase, and equalizer
switching. The logic signals are produced for each of the individual signal
channels.
The switching logic is made up of four similar circuit sections, one
section for each of the four signal channels. ( See Fig. 14 -7 for a functional
block diagram of the switching logic.) Only the section for channel A is
described. Diodes X6 and X7 form an AND gate receiving the record/
playback command and the EAc signal. The AND -gate output causes driver
Q2 to go into conduction whenever both gate inputs are high. The output
from the collector of Q2 is routed to pins 3 and C of J12. Here the signal
is split into two paths, one going to the record -amplifier channel -A record
gate, and the other going to the channel -B head -preamplifier erase -gate
input. This circuit is always inhibited when the HS -100 is not in the record
mode.
Diodes X3 and X4 form an AND gate receiving the logic EAc signal and
the inverted record /playback command. This gate, whenever both inputs
are high, causes driver Q3, Q4 to go into conduction; the channel-A
equalizer gate signal is generated at the collector of Q3. The collector output
from Q3 is connected to pins 2 and B of J12 and is supplied to the rf
equalizer board in the esu. It should be noted that this AND gate is inhibited
whenever the record mode is selected.
The record /playback command is also applied to relay driver Q14.
Selection of the record mode shifts the record /playback command level to
+3.8 volts and drives Q14 into conduction, providing a current path from
+28 volts to ground through the coil of K1. The energized relay switches
+12 volts from pin 15 to pin Y of J12 to provide the +12 -volt record
signal to the record amplifier. It also switches +28 volts from pins 22 and
Z to pins 20 and X of J12 for the head relays on the head preamplifier
assemblies.
003 NI
003 Ng
093 Pd
]Y3 Pd
093 Nd
003 Pd
N Nd
003 od
003 Pd
093 Pd
Fig. 14 -6. Functional block

OA
31.
diagram of HS -100 signal system.

RF Equalizer
The rf equalizer operates on much the same principle as that of con-
ventional vtr's. Its purpose is to restore the original carrier -to- sideband
relationship of the recorded information. Two cosine equalizers are used.
One is an overall master equalizer, and the other has controls for each
individual channel.
Equalizer Gate Signal
+4V AOV PB +0.7V
0511
X3, X. )-o. Driver PB -12V-
J12 LB Ch A
Q3, Q4
2 Eq Gate
FACI
X6 X7
Driver Rec+0.9V Rec-+12V
Q2 1
Record A
1 Erase B
X8. X9
Driver
Q5, 07
LF (ChB
6 { Eq Gate
Driver
06 ( Record B
( Erase C
Driver Ch
X13, X14
Q8, Q9 D Eq
C
Gate
tCC L5 X16. X1
Driver
010 a Record
Erase D
C
018.X19 Driver
o11, 012 n Ch D
Eq Gate
Driver Record D
X21. X22
Q13 Erase A
PB+4V
15 +12V
PB +1V
Rec /PB I
Y +12V to Pin 1
of Record Amp
Command( Gate
22 +28V
Switch
20 l +28 V to Head
X ( Relays
Rec 3.8V Kl (Energized During

PB 0.7V Relay Record)
Driver
.11
014 +28V
112
Note: Voltage levels indicated are approximate values only.
PB - Playback or Reproduce It Not in Playback)
Rec Record 1 Not in Record)
Fig. 14 -7. Functional block diagram of switching logic.
Demodulator Module
The demodulator is similar to that of conventional vtr's. The function is
to limit the rf signal and detect zero crossovers of the fm carrier. Pulses are
then formed by a tunnel diode and associated switching circuitry to produce
narrow rectangular pulses, one for each carrier crossover point (therefore
at a frequency of twice the carrier frequency) These pulses have short

.
rise times and are approximately 25 nanoseconds in width and 2 volts in

peak -to -peak amplitude. The pulses are then fed to a low -pass filter on
another module.
30-MHz Modulator
The 30 -MHz modulator assembly contains a 30 -MHz oscillator, an
amplitude modulator, and a simulating network. Fig. 14 -8 is a functional
block diagram of the 30 -MHz modulator.
Transistor QI, crystal Yl, and additional LC components form a 30 -MHz
oscillator. Crystal YI functions in a series -mode overtone operation. The
oscillator output passes through buffer amplifier Q2 and carrier -level con-
trol R8 to current- source amplifier Q3, Q4. The carrier signal from the
common collectors of Q3 and Q4 is connected to diodes X1 and X2. The
amplitude- modulated signal enters complementing amplifier Q5, Q6 at a
high impedance. The output from Q5 and Q6 is a low -impedance circuit
split into two signal routes. The first route is through output pin 8, into
the external half -line delay line, and to the 30 -MHz delay -line switcher
and demodulator board. The second route is through the delayed /undelayed
signal balance control, R22. The arm of control R22 routes the signal into
a network simulating the amplitude characteristics of the delay line in use.
Some of the network components are factory selected to match the delay
line in use. However, in some equipment this network may not be required.
Terminals on the board allow for signal routing through the network or
around the network, as required. The network output is routed through pin
I of J313 to the 30 -MHz delay -line switcher and demodulator board.
The output from the low -pass filter enters the module at pin 22 of J313
as a carrier -free demodulated video signal, which is routed to video-level
control R48. The arm of R48 routes the signal to amplifier QI1, which is
supplied by constant -voltage source Q12. A capacitor in the emitter circuit
of Q11 allows for peaking of the amplifier to obtain the desired frequency
response. The output from Q11 is coupled through C34 to phase splitter
Q9, which is a dual transistor. The emitters of Q9 are supplied by constant -
current source Q10. The two outputs from the collectors of Q9 are 180°
out of phase and are coupled capacitively to the bases of complementary
emitter followers Q7 and Q8. Control R34 in the base circuits of Q7 and
Q8 provides a balance adjustment for the second harmonic of the carrier.
The two outputs from complementary emitter followers Q7 and Q8 are
applied to modulator diodes X1 and X2. The voltage level at the anode of
X1 is approximately -0.7 volt, and the level at the cathode of X2 is
approximately +0.7 volt. These voltages establish a back bias on the diodes.
Delay Line
The delay line employed between the 30 -MHz modulator and the 30-
MHz delay -line switcher and demodulator is of the ultrasonic glass type,
O OL
ov,
O w
a
fI,
Y w á 0-X= VX
0-0
-
O
NA A.H
-vV91 N
á0
tc,o.
L
aN 0d
Fig. 14 -8. Functional block diagram of 30 -MHz modulator.

having integral ceramic transducers. The delay line operates at a center

frequency of 30 MHz and provides a delay of 1/2 line of video ( 31.798 µs
-!-10 ¡Ls). The output from the delay line is applied to one input of the
30 -MHz delay -line switcher and demodulator.
30 -MHz Delay -Line Switcher and Demodulator

The 30 -MHz delay -line switcher and demodulator assembly contains the
switching gate, the gate drivers, the demodulator, and associated amplifiers.
No delay-line switching occurs during the record mode of operation. Fig.
14 -9 is a functional block diagram of the assembly.
Switching Gate-The switching gate is made up of three diode assem-
blies, X1 through X3, and each assembly contains four diodes. The un-
delayed and delayed modulated video signals are applied as inputs to the
gate. Only one of the two signals is allowed to be gated through into the
demodulator section at any given time.
Gate Drivers -The diode gate is driven by gate drivers, under control
from a logic switching signal. The logic signal (R) enters the board at
pin 9 and is applied to two gate- driver circuits. The input signal is binary
1 when a positive signal of 3.5 to 5.0 volts appears at pin 9. This positive
signal switches on Q10, which in turn switches Q8 and Q9 off. The voltage
at the emitter junctions of Q8 and Q9 will then become negative 5.6 volts.
The emitter output is series connected with a control, which controls the
current through the diodes and effectively controls the rise time of the
switching waveform. This emitter output is connected to six diodes in
assemblies X1 through X3 through resistors R28 and R31. The input
signal from pin 9 also switches on Q11, which in turn switches on Q14,
causing Q15 to switch on and Q16 to switch off. The resulting positive 5.6
volts at the emitter junction of Q15 and Q16 is applied through resistors
R29 and R30 to the six other diodes in X1 through X3. The negative
voltage applied through resistors R28 and R31 causes diodes X1D, X2D,
and X3B to be forward biased. The positive voltage applied through resis-
tors R29 and R30 causes forward biasing of diodes X1A, X2C, and X3C.
Therefore, the undelayed signal from pin 4 will be gated through and
appear at coupling capacitor C22. Diodes X1B, X1C, X2A, X2B, X3A, and
X3D are back -biased and block the delayed signal from pin 7. Any spurious
signal appearing from pin 7 through X3A and X3D finds its path blocked
by X2A and X2B; furthermore, diodes X3B and X3C are conducting and
effectively ground the junctions between the X2 and X3 diodes.
When the logic -signal input at pin 9 becomes less than +3.5 volts, or
inhibitive (binary 0) , the transistors in both gate drivers reverse their
states. The signal into resistors R29 and R30 becomes negative, and the
signal into resistors R28 and R31 becomes positive. This causes a polarity
reversal of all diode biases, shutting off the path from pin 4 through the
gate and switching on the path from pin 7 through the gate. The input into
the bandpass amplifier is then from the delay line.
0 >5
o
ó
O
5
5d
O -
1 ï
r- CC
K K
r
X X
L_
i.
N
o M
Fig. 14 -9. Functional block diagram of 30 -MHz delay -line switches

and demodulator in HS -100 system.
Demodulator-The bandpass amplifier (Q6 and Q7) is tuned to provide

a second -order Butterworth response and drives emitter follower Q5. This
buffer drives current source Q3, Q4, which has its output connected to
detector Q1, Q2. Coil L3 in the detector input circuit removes the dc com-
ponent of the demodulated signal and eliminates the effect of stray capaci-
tances at the carrier frequency. The detected video signal is routed from
the collector of Q1 through a filter to emitter follower Q12. The output
from the Q12 emitter is routed through a second filter and output -level
control R32 to emitter follower Q13 and pin 19. The signal at pin 19 is
the demodulated video, timed to specific requirements of the selected
reproduce mode.
Output Video Module
The output video assembly receives the demodulated video signal from
the demodulator. The video signal is amplified, fed through a de- emphasis
network (external to this module) , and further amplified before it appears
as an output. This signal is then routed to the Amtec unit in the opu.
Chroma Inverter Module

The chroma inverter assembly contains a low -pass filter, a bandpass filter,
a chroma inverter, amplifier stages, and the logic circuits controlling the
inverter. Fig. 14 -10 is a functional block diagram of the chroma inverter.
Chroma Phare Inverter -The composite video, processed by the Amtec
unit, enters the board through pin 4 of J315 and follows two signal rout-
ings. One route is into amplifier Q1 and a low -pass filter; the other is
directly into a bandpass filter. The video signal from the collector of Q1
is routed through the low -pass filter, which passes only the luminance
portion of the signal. The path through the bandpass filter allows the sub -
carrier with its sidebands to pass and couples this signal to a driver stage
composed of Q9, Q10, and Q11. The output from Q10 and Q11 is a
low- impedance signal feeding drive transformer Tl. This transformer
converts the single -phase signal into two signals of opposite phase. The
two opposite -phase signals are applied to the diode gate composed of Xl,
R6, and R38. Depending on the polarity of the gate-driver signal, the
gate will pass the signal through one of two pairs of diodes to gate trans-
former T2. The gate- driver signal from the collector of Q12 is approxi-
mately +6 volts for the inverting condition, in which the diodes effectively
cross -connect T1 and T2. No inversion takes place when the gate-driver
signal is approximately -6 volts, since the diodes then connect corre-
sponding ends of the windings of T1 and T2 together. The single -phase
output from T2 is applied to emitter follower Q8. Control R38 in the
diode gate is set to obtain chroma balance between the noninverted and
inverted chroma signals. Control R23 in the base circuit of Q8 is factory
set to the correct bias level for Q8. The output from Q8 is applied to the
anode of zener diode X2.
CE
s VI
E o
Fig. 14 -10. Functional block diagram of chroma inverter.

The output from the low -pass filter ( luminance signal only) passes
through amplifier Q2. The luminance gain is factory set to 1:1 by a con-
trol in the base circuit of Q4. Chroma and luminance information are
combined by the output adder consisting of X2 and Q2. The recombined
signal is coupled through amplifier Q3, Q4, Q7 and complementary emit-
ter followers Q5 and Q6. The output from Q5 and Q6 provides a low -
impedance signal to pin 6 of J315. At this point, the signal is a standard
NTSC color signal ready for further processing by the Colortec and the
processing amplifier.
Signal CH from the logic circuitry on this board controls the chroma -
inverter function. This signal is binary 1 if the chroma inverter is required
to provide a shift of 180° in the phase of the chroma subcarrier; it is
binary 0 for any other condition. Signal CH is applied to the base of Q13.
Q13 conducts if CH is 1, and the resulting collector potential of Q13
drives Q12 into conduction. Transistor Q12 is tied between positive and
negative 12 volts. A zener diode in the emitter circuit of Q12 limits the
collector potential of Q12 to approximately +6 volts when Q12 conducts.
Transistor Q13 is held in cutoff if CH is 0, and the resulting collector
potential of Q13 holds Q12 in cutoff. Transistor Q12 is not conducting,
and the collector potential of Q12 is at approximately -6 volts. Thus, the
polarity of the gate -drive signal from Q12 changes with the logic state
of signal CH and controls the chroma -inverter circuit.
Chroma - Inverter Logic -Logic circuitry contained on the choma- inverter
circuit board controls the switch -in and switch -out of the 180° phase -
reversing network of the chroma inverter. See Fig. 14 -11 for a functional
block diagram of the chroma -inverter logic circuit. (Also refer to the
function -designators table in Section 14 -5 for the meanings of some of
the abbreviations used in the following descriptions.)
Since the HS -100 operates in the E -E mode whenever a record or fast -
search mode is selected, the output from the chroma -inverter logic is in-
hibited during the record and fast - search modes. Two signals, 8 and P4
enter the board at pins 11 and 16 of J315 and are connected to NAND gate
A4A. Either of these two signals can thus inhibit the CH output of A4A.
Three operational conditions of the chroma inverter provide for correct
chroma phase switching during (1) forward reproduce, (2) reverse re-
produce, and (3) alternate -field reproduce. Each of the three conditions
includes normal motion, slow motion, and freeze motion.
Forward Reproduce- (Refer to Figs. 14 -11 and 14 -12.) During for-
ward reproduce, the input at pin 18 of J315 is binary 1. This signal (K) is
inverted by A5B, and the resulting signal goes to NAND gate A2A, caus-
ing this NAND gate to be inhibited. The output of A2A is, therefore, bi-
nary 1 for forward motion. The normal/alternate -field record switch in
the esu is in the normal position and selects R' for application to pin 20
of J315. This signal is then applied to the PK input of JK flip -flop A3
and, through inverter A5D, to the PJ input of the same flip -flop. Pin 7
O 10 Id
i
O
á â
U
a.IP ^á Q
h1 =1 0 8 0
m
$
Fig. 14 -11. Functional block diagram of chroma -inverter logic.

io
°
--. 9.
o.
O ._....
ó.
-. o°o _ °
ó 3v
co
c óP
N O
---. -.- O a G
Ñ V
Ó a0 ! p S
.
-..' ó
3
o
_
:
G
t
íJ
ÿ
U
O
"Z t_
91 .s É
p n
°oñ ._,
-ú
° ° ó
SO
-
Zi Á?.
m
5
m L"- á
OP r ti
O 6
- N G_
o+
% Ó Ó
L Lg
Q CC <J U
.. N re;
15
-
E
ó
- Oo
_
..
2
-. o°
°o
r _ 5
O d-
-
-rr
Z =Q L
Fig. 14 -12. Chroma -inverter waveforms, forward reproduce mode.

of A3 receives the inverted C -pulse input from pin 19 of J315 through

inverter A5C. Thus, signal C becomes the clock pulse for flip -flop A3.
Signal R' is binary 0 when the 1/2-line delay line is out and sets flip-flop
A3, resulting in an output of binary 0 on pin 12 of A3. This output is
R',,, which changes its level whenever the 1/2-line delay is switched in or
out. Switching in of the 1/2-line delay causes R'e to change from the
binary 0 level to the binary 1 level, and vice versa. Signal R'(. is routed
through inverter A2C and is then differentiated by C39 and R44. The
differentiated Re' is fed to pin 13 of NAND gate A2B.
Pin 12 of NAND gate A2B is at the binary 1 level during forward mo-
tion, since its signal comes from pin 2 of NAND gate A2A. Pin 13 is held
normally at a level equivalent to binary 1 by the 4.5 volts supplied through
resistor R44. The output at pin 14 of A2B is, therefore, equal to a binary 0.
With these levels established, it is apparent that only a negative pulse from
differentiating network C39 -R44 will cause a change in the NAND -gate out-
put. A negative pulse occurs only when R' changes from binary 0 to binary
1, which is equivalent to the switching in of the 1/2-line delay. The positive
output pulse at pin 14 of A2B is wired to the clock input (pin 7) of JK
flip -flop Al. Flip -flop Al complements its own output for every clock
pulse applied. Output pin 12 of the flip -flop is wired to pin 10 of NAND
gate A4A and represents signal U. Signals P4 and B also are applied to
NAND gate A4A. The output from pin 14 of this gate is fed to another
NAND gate, A4B. Since all other inputs to this gate are open -circuited, it
acts as an inverter; its output is CH, the chroma- inverter control.
Signal CH follows the Boolean equation:
CH= UP., B
Thus, CH will change its state with every change of U, whenever the HS-
100 is in reproduce and not in a fast search. Since U is a function of R',C
is also a function of R', and CH changes every time that the 1/2-line delay
is switched in. The CH waveform, therefore, is a divided -by -2 version of
the R' waveform. Signal CH is routed to the base of Q13 (Fig. 14 -10) and
to pin 13 of J315 (Fig. 14 -11) .
The condition of the chroma inverter is changed every time that the
1/2-line delay signal (R') changes from binary 0 to binary 1. Changing
of R' from binary 1 to binary 0 has no effect on the chroma- inverter state
during the forward reproduce mode.
Reverse Reproduce- (Refer to Figs. 4 -11 and 14 -13.) The only dif-
ference between the forward and reverse reproduce modes is the signal
at input pin 18. This signal is binary 0 during reverse; therefore, it does
not inhibit NAND gate A2A, and Jc pulses can now pass through this
gate. The output from this gate is Jc', which passes through NAND gate
A2B and complements output U of flip-flop Al for every Je pulse. Signal
CH changes with every change of U and results in changing of the chroma
phase. However, during slow-motion reverse, the R' waveform must also
J C Ó
V
5
L L
Q K U
Ó a0
Ó
I 5o Na, _ 5
cc Z Z
Fig. 14 -13. Chroma -inverter waveforms, reverse reproduce mode.

be considered. This waveform exercises its control in exactly the same

manner as it does in forward motion, and U is, therefore, changed every
time the 1/2-line delay is switched in.
Signal C11 changes from binary 1 to binary 0, or vice versa, every time
that a Je pulse occurs or the 1/2-line delay switches in. The Je pulses occur
whenever the head signal carries a new field from the discs. No J, pulses
occur when the head signal is a repetition of a single field, as is the case
during slow motion or freeze.
Alternate -Field Reproduce- (Refer to Figs. 14 -11 and 14 -14.) The
input to pins 18 and 20 of J315 is changed by the alternate -field record/
normal switch in the esu. The input to pin 18 is changed from signal K
to ground potential, and the input to pin 20 is changed from R' to B0.
Flip -flop A3, followed by inverter A2C and differentiator C39 -R44, now
produces clock pulses for flip-flop Al which are derived from Br and re-
timed at C. Flip -flop Al also receives clock pulses derived from Je, since
the controlling input at pin 18 of J315 is now at ground potential. This
ground potential is the level of binary 0 as in the reverse reproduce mode.
However, it should be noted that the Je inverter, A5A, has capacitor C38
connected between the output and ground. Capacitor C38 and internal
components of A5A form a delay network producing a delay of 20 micro-
seconds. This delay of waveform Je is of no importance in modes other
than the alternate -field reproduce mode. A delayed Je pulse may comple-
ment flip-flop Al about 20 microseconds after a BG- derived clock pulse
changes the state of flip -flop Al, and thus re- establishes the state of flip -
flop Al which was present prior to the BG- derived clock pulse. The end
effect of this coincidence control is to prevent a change in the chroma
inverter if (and only if) the 1/2-line delay switches in during the pres-
ence of a Je pulse. The chroma inverter will change whenever the 1/2-line
delay switches in during the absence of a Je pulse, and whenever the I/2-
line delay switches out during the presence of a Je pulse.
14 -5. CONTROL LOGIC SYSTEM
The HS -100 system is totally dependent on "computer- type" circuitry,

and as stated several times previously in this text, it is mandatory that
the technicians concerned with modern video equipment have a good work-
ing knowledge of logic theory. Such a background is assumed in this de-
scription. The instruction books for this system are complete (and large) ,
with complete schematics presented. Space in this text permits only a basic
description, sufficient to familiarize the reader with fundamentals of
slow-motion, stop- action disc recording logic functions.
Control Functions
The control logic receives its commands from the operator through the
control -unit push buttons and generates waveforms for the control of:
00
o
a
O.
m
---
o É
ó -
o
El.
O.
o
Q mU U
Fig. 14 -14. Chroma -inverter waveforms, alternate -field reproduce mode.

(A) Head stepping

(B) Record signal switching
(C) Playback signal switching
(D) Half -line delay switching
(E) Chroma -inverter switching
Logic circuit boards are located in the control unit, electronics signal
unit (esu), and disc servo unit (dsu). The control -unit logic responds
to the push- button commands by generating basic command signals (re-
cord /reproduce, forward /reverse, alternate-field /continuous, slow- motion/
normal, fast forward, fast reverse) and by generating the slow- motion
speed -control waveform. In addition, the control -unit logic board contains
electronics to drive the indicator clock and cue needle, lamp- indicator
drivers, and a 4.5 -volt power regulator.
The switching -logic module located in the esu generates the switching
waveforms for the record, erase, and playback signals. Also generated are
the control signals for the half-line delay and chroma- inverter circuits.
The clock -generator circuit, which times these waveforms, is locked to
incoming reference sync by signals generated in the sync -generator module.
Supplementary timing information is received from the slow- motion
oscillator in the control unit during slow- motion playback and from the
fast -search logic module during the fast - search mode. In the alternate -
field mode, timing information is received from the alternate -field logic
circuit located on the fast - search logic module.
The logic which controls the stepping of the head carriages is located in
the dsu on the carriage logic Al and carriage logic A2 modules.
For description and analysis purposes, the logic circuits in the control
unit, esu, and dsu are shown in block diagram form in Figs. 14 -16, 14 -17,
and 14 -25. Each block represents a functional circuit element of the system
and is identified by name. Each logic input or output is designated by a
letter which indicates its function. These function designators, plus the
other secondary functions, are listed and defined in Table 14 -5.
Push- Button Logic-Nine of the twelve push- button switches on the
control unit are interlocked with flip -flop circuits. See Fig. 14 -15 for a typi-
cal circuit arrangement. Pressing one of the nine buttons causes the asso-
ciated holding flip -flop to be set, and the active output of the flip-flop
is utilized in the control -unit logic. Depression of the push button momen-
tarily transfers the switch output from the normally closed contact to the
normally open contact, resulting in a voltage -level change. Application of
a 0 -volt level (binary 0) sets the flip-flop. The duration of the applied
0 -volt level is immaterial. The return to a positive 4.5 -volt (binary 1) has
no effect on the flip-flop state. Only a binary 0 at a reset input of the flip -
flop can cause a change of state to the reset condition. The push- button
initiated signal which sets one flip-flop may be routed to the reset inputs
of other flip -flops, causing these flip -flops to be reset. This electronic inter-
Table 14 -5. Function Designators
Designator Function Designated
A Slow- motion control waveform output from control unit.

A, Output from freeze button. In freeze mode Al =1.
Aa Output from single -frame advance button. Normally A3 = 1.
A, Output from alternate-field logic.

AF Output from alternate-field switch.
BG Output from clock generator.
B' Internal connection in alternate -field logic.
C Clock -pulse output from clock generator.
CH Output from chroma- inverter logic.
DG Output from slow-motion logic. Equals Be in normal and Zr: in slow -
motion mode.
EAC Positive head -switching pulses for channels A, B, C, and D, respec-
ESC tively. Suffix "C" means that the pulses are timed by the "C" clock
Enc pulse. Pulses are output from the head -retiming logic.
Enc
EA,. Timing pulses for channels A, B, C, and D, respectively. Suffix "G"
Esa means that the pulses are timed by the "G" clock pulses. Pulses are
Ecc output from the head logic.
EDc
EA,: Head -switching logic output from the reverse- motion logic. Suffix
ECK "K" indicates the functions are subject to forward /reverse control
function K.
F Field identification pulse. F = 1 when a horizontal -line pulse occurs
at the same time as the first vertical pulse during equalization (even
field).
FAG Carriage-stepping logic pulses for channels A, B, C, and D, respec-
FaG tively. Suffix "G" means that the pulses are timed by G clock pulses.
Fcr. Pulses are output from the carriage logic.
FDG
FAC Carriage- stepping logic pulses for channels A, B, C, and D, respec-
Fac tively. Suffix "C" means that the pulses are timed by C clock pulses.
Fcc Pulses are output from the carriage -retiming logic.
FDc
FAK Carriage- stepping logic pulses for channels A and C. Suffix "K"
FCK means that the pulses are timed by K. Pulses are output from the
reverse -motion logic.
FA' Carriage-stepping logic pulses for channels A, B, C, and D, respec-
Fá tively. Pulses are output from the carriage control logic.
Fc'
FD'
Feco Carriage- stepping logic pulses for channels A, B, C, and D, respec-
FBCO tively. Pulses are output from the error -correction logic. Suffix "O"
Fcco indicates that a pulse on this line will cause carriage to move out
Foco from the disc center.
FACI Carriage- stepping logic pulses for channels A, B, C, and D, respec-
For-, tively. Pulses are output from the error-correction logic. Suffix "I"
F,. indicates that a pulse on this line will cause carriage to move in
Fnr, toward disc center.
FF Fast -forward search command. Originates in control unit.
Fe Fast -reverse search command. Originates in control unit.
(Table continued on page 518.)
Table 14 -5. Function Designators (Cont)

Fs F function that has been modified by the fast -search logic circuitry.
G Output from clock -pulse generator. A preclock pulse coincident with Ts.
J Slow- motion clock pulse. Output from slow -motion logic.
Je Slow- motion clock pulse timed by C.
K Forward/reverse- motion logic control. Output from reverse -motion
logic.
K' Output from alternate-field switch.
L Internal connection in sync separator.
L'
M Output from carriage- reversing logic. M = when carriages move
1
out; M = 0 when carriages move in.

N Output from reverse -motion logic.
P, Output from normal push button on control unit and is the unquan-
tized normal /slow command. P, = in normal mode.
1
P_ Output from forward /reverse push button on control unit and is the
unquantized forward /reverse command. P2 = in forward. 1
P2 Output from control logic in control unit. P3 = in alternate -field

1
record mode; otherwise Pa = O.

P4 Output from control logic in control unit. Unquantized record /repro-
duce command. P1 = 1 in record mode.
Q Output from monitor carriage pulses switch on dsu. Q = 0 when
pressed to enable carriage monitor lamps.
R Half-line delay switching waveform. Output from alternate -field logic.
R = 1 when delay is out.
R' Same as R but output is from the half -line logic.
RA Output from alternate-field switch on the esu.
SR Servo reference -pulse output from sync separator. Pulse is positive
coincident with the first serration of composite vertical sync.
Sau Identical to SR except delayed 15 As when in record mode.
Ss A 5 -µs output pulse from the sync separator occurring at a horizontal -
line rate.
SYD A 5 -µs output pulse from the horizontal delay. Delayed 63.5 µs.
T Output from sync separator. Master timing signal of 9 horizontal -line
duration occurring at a vertical rate.
is Identical to T in all modes except fast -search modes. In fast search,
T3 is a 450-µs pulse occurring 4.5 times faster than vertical rate.
TsD Output of T -pulse delay.

U Internal connections in alternate -field logic.
V
W Output from control -unit logic. W = in slow motion.
1
Ws Identical to W except in fast -search modes. Ws = 0 in fast search.

Xe Output from outer carriage -limit switches on channels A, B, C, and D,
Xn respectively. Command = when end stop is reached.
1
Xc
XD
Ye Output from inner carriage -limit switches on channels A, B, C, and D,
YR respectively.
Ye
YD
Z.: Output from slow- motion quantizer. A preclocked form of A, the
slow- motion control waveform.
Table 14 -5. Function Designators (Cont)

a Internal connection in fast -search logic.

ß Output of alternate -field logic.
0 Internal connection in fast -search logic.
C, Delayed clock pulse generated in the posicast timing module. Pulse
is 3 -µs long with front edge coincident with the trailing edge of C
pulse.
lock of the flip -flops serves to terminate one mode of operation when
another mode is selected. A change from normal reproduce to slow mo-
tion, for example, terminates the normal mode (resets the normal flip-
flop) and initiates the slow- motion mode (sets the slow-motion flip-flop) .
For the following descriptions of push- button functions, refer to the
block diagram of Fig. 14 -16.
Fast- Forward and Fast- Reverse Controls (FF and FR) -The fast -for-
ward switch (FF) and the fast -reserve switch (FR) are used without flip -
flop circuits. Both circuits are similar and therefore only the fast -forward
circuit is described. Pressing the fast- forward switch removes the normally
present ground potential from the input of an inverter and changes the
inverter input to positive 4.5 volts. Inverter output equals binary 0 1
during fast search and equals binary 1 during the absence of a fast - search
mode. The fast - forward and fast -reverse command signals, FF and are P,
routed into two paths. The first path is to the control -unit connector for
routing to the esu, and the second path is to a NAND gate. The output
from this gate is binary 1 when the freeze, fast- forward, or fast -reverse
push button is pressed. The output is inverted and fed to the reset input
of the record, forward, and reverse holding flip-flops, as well as the set
+ V1
Push- Button Switch
Function 1 3
Push -Button +4.5V
Indicator
A
101(
15
Reset Path to R57 2 470
Other Flip -Flops
Set 6 0101
012
R101
Q 500
To Logici illumination
Level
k1b Clamp
X21
13
10
Q +28V
Inverter Reset I
To Logic)
Inputs I
11 12
Fig. 14 -15. Simplified diagram of typical push- button circuit.

i ó j
ILL
u
UU
r-
0.11 0
U
Fig. 14 -16. Logic block diagram of control unit.

input of the freeze flip -flop. Thus the termination of a fast -search mode
always leaves the system in the freeze mode.
Frame- Advance Control (FR ADV) -The frame -advance switch op-
erates without a flip -flop, and the switch output is normally at ground
potential. Pressing of the switch results in a change of its output to posi-
tive 4.5 volts; the output reverts to ground potential on release of the push
button. The output from the switch is designated as signal R1 and is
routed to the esu and to the control processing logic.
Cue Control Circuit-The cue control switch is connected to the clock
input of a JK flip -flop. Pressing the cue push button changes the posi-
tive 4.5 -volt clock input level to ground potential. Successive pressings
of the cue push button cause the flip -flop to set and reset. The flip-flop is
reset to obtain a freezing of the cue pointer, and it is set when the cue
pointer is slaved to the timer hand. Remote cue is obtained by producing
the level change from 4.5 volts to ground at a remote point and routing
this level through the normally closed contacts of the switch. Since the
flip -flop state is changed at each level change, a cue may be set at the
control unit and cancelled from the remote point, or vice versa.
Slow-Motion Oscillator-Three slow- motion switches (SM 1, SM2, and
SM3) provide two fixed and one continuously variable motion speeds.
The slow- motion circuits use a unijunction- transistor relaxation oscillator
to produce slow -motion waveform Z. A change in oscillator frequency is
obtained by controlling the resistance value in the RC section. Fixed slow-
motion speeds are provided by switching in preset resistors, and variable
slow- motion speed is provided by switching in a variable resistor oper-
ated by the speed-control lever.
The slow-motion oscillator output (Z) is connected to the control
processing logic circuitry, where it is combined with the freeze waveform
(A1) and the frame -advance waveform (R1) before being sent to the
switching logic module as the slow- motion reference waveform (A) .
Control Processing Logic -The outputs of the push -button holding

flip -flops and the slow- motion oscillator are combined in the control
processing logic circuitry to produce the basic command signals which
(along with FF and FR) control the recorder. These are:
(A) Forward/reverse command (P2)
(B) Record /reproduce command (P4)
(C) Alternate-field record command (P3)
(D) Slow -motion reference waveform (A)
(E) Slow- motion /normal command (W)
Clock MDA -The clock MDA drives the timing- indicator clock in
synchronization with the carriage stepping. The input signals are
(from the esu), which goes to binary 0 each time head D is recording or
reproducing, and signal K, the quantized forward /reverse command (from
the dsu) , which indicates the direction of clock motion.
The indicator clock is driven by a stepper motor having two pairs of

bifilar windings. One pair is used for stepping in the forward direction,
and the other pair is used for reverse stepping. Stepping in either direc-
tion is accomplished by alternately applying pulses to first one side and
then the other side of the winding.
Electronics Signal Unit Logic
Refer to the logic block diagram of the esu, Fig. 14 -17, during the
following basic description of this logic.
Sync- Separator Module-The sync -separator module receives the com-
posite sync and performs the processing required to obtain pulse signals
suitable for control of the HS-100 system. Refer to Fig. 14 -18 for the rele-
vant waveforms in this module. These waveforms are described in para-
graphs that follow.
Four primary pulse signals are derived from the composite sync and
are assigned the symbols Sß, Sr, F, and T. The SR, or servo reference,
pulse occurs at the trailing edge of the first vertical -sync pulse. The S,
pulses are derived from horizontal -line pulses and occur at the horizontal
frequency. A T -, or vertical -interval, pulse embraces the time interval from
the first to the last equalizing pulses, including the vertical -sync pulses.
The F, or field- identification, pulses identify the beginning of odd fields
and are absent at the beginning of an even field. Two secondary pulse
signals, the L and L' pulses, are derived in addition to the other pulses
for utilization within the sync- separator circuitry.
Servo Reference Pulse (SR) -The servo reference pulse is obtained in
a vertical -sync separator by gating through the first serration of the verti-
cal -sync pulse.
Line Sync Pulse (Si-) -The line pulses (Si) are generated by a hori-
zontal-sync separator consisting of two monostables. The first monostable
triggers on the negative -going leading edges of pulses and produces out-
put pulses of 45 microseconds duration. Since the period of 45 micro-
seconds is longer than one -half line period but shorter than a whole line
period, and the monostable cannot be retriggered until it resets, the result-
ing output represents every line pulse plus alternate equalizing pulses
and alternate vertical-sync pulses. Therefore, the output is a succession of
pulses at line frequency, with the vertical -sync information effectively
removed.
Field Identification Pulse (F) -The field- identification pulse (F) is
generated by gating the line sync pulse coincident with the first vertical -
sync serration pulse. The Boolean equation for this pulse is:
F =LS1
The 17-microsecond L pulse generated in the SR processing circuitry is
gated against Sy, and a NAND gate produces F. The F pulses occur only
at the transitions from even to odd fields.
Control Unit
Sync Input P4 P3 A W FF Fe AF
11111.111111j
Servo Ref SRD To Servo
Delay
Sync
Separator
Horiz Ref
To Amtec & Colortec
Delay S YD
T Pulse
Delay
TSD
WS
I [
E -E
0 Command
yy Fas
Sear h Fs Clock Slav Motion
FF Logic Generator
G
Logic
Bi
Y.-.
P25
DG
3
Delay Line Slav Motion
4 Alternate Logic Quantizer
Field
Logic
f As
To Record and
BG BG
Playback Switchers
All Field I
tttt
_
Normaall't 4.5
AF
DG EAC EBC ECCEDC
? Gnd
Head
EaG EAK Head
E EBG Retiming
Logic bG
ECK Logic
EcG
EdG EDG
L
As Ra K'
PG
Chroma
Inverter
i
¡ 7
K P2S xaa aa EAG I EBGÍ ECCI EDG EAK ECK
DisdServo Unit
Fig. 14 -17. Logic block diagram of electronics signal unit.

Fig. 14 -18. Sync- separator waveforms.

Vertical Reference Pulse (T) -The vertical pulse (T) is a positive -

going return -to-zero pulse that begins at the end of the last line -sync
pulse, extends through the equalization and vertical-sync pulses, and ter-
minates prior to the first line -sync pulse of the next field. The leading
edge of the T -pulse is generated from Sy by a flywheel oscillator and a
ripple-through counter. Transistors with their associated components form
a free -running multivibrator. A starting circuit assures correct starting of
the multivibrator on initial power application. Normally the multivibrator
is pretriggered by the incoming Si pulses. However, should one or more
of the line -sync pulses be missing, the multivibrator will continue to run
at its natural frequency, which is determined by the setting of a calibra-
tion control. The multivibrator is set at a frequency approximately 5 per-
cent lower than the horizontal frequency, allowing for insertion of up to
20 line pulses without error in the total number of sync pulses per field
in case of line -sync dropouts in the Sy signal. The output is differentiated
and inverted and used as a clock pulse to trigger a JK flip -flop.
The JK flip-flop is the input stage of a chain of ten JK flip -flops con-
nected as a ripple- through counter. The second input to this counter is
the L' pulse (a 47- microsecond pulse) wired externally to the board. The
L' pulse is inverted and applied to the J inputs of the flip -flops as L'. After
a second inversion, signal L' is applied to all Ps inputs of the flip -flops. The
ripple- through counter must count equal numbers of pulses for odd and
even fields. It also must start its counting exactly 258 line pulses prior to
the first equalizing pulse. To meet this constraint, the counter must be
reset after the second vertical-serration pulse, and this is accomplished
by the L' signal.
When the 258th sync pulse has been counted, the output of the last
flip -flop, through an inverter, changes the state of the NAND -gate flip -
flop generating the leading edge of the T- pulse. This differentiated signal
is also routed to an inverter followed by a second inverter, and the output
of both inverters is applied to a four -stage chain counter to preset all of
its four JK flip -flops. These flip -flops form the second counter, which re-
ceives a signal derived by gating processed sync with the 600 -microsecond
pulse starting at the first vertical -pulse serration. Twelve pulses into the
counter cause the state of the output flip-flop to change in coincidence
with the end of vertical equalization. The output is inverted, differentiated,
and applied to a NAND -gate flip-flop producing the trailing edge of the
T- pulse.
Fast - Search Logic Module-The fast -search logic module is located in
the esu. The module produces pulses and signals required during fast -
search modes. Its circuitry replaces the normal -motion T- pulses with fast -
search Ts- pulses of increased repetition rate. Other functions generated
are the O command, which directs the electronics into the E -E mode; the
fast forward /reverse command (P25) , with inhibition of the normal
forward /reverse command (P2) inhibition of the field- identification
;
pulse (F) ; and inhibition of the half -line delay command. Fig. 14 -19 is
a timing diagram showing the relationships of relevant waveforms.

Switching Logic Module -The switching logic module is located in the
esu (Fig. 14-17 ) The module contains circuit functions which include
.
the clock generator, the slow- motion quantizer, the slow-motion logic,
the head logic, the head retiming logic, and the half -line delay logic. Each
of these functions is described below; timing diagrams show the rela-
tionships of the waveforms.
Clock Generator -The clock generator is supplied with input signals
Ts and Fs ( Fig. 14 -20) Both input pulses originate at the fast -search
.
logic board. From the two input pulses, the clock generator produces three
basic timing pulses, which are required to synchronize the switching of
the system logic. The preclock pulse (G) , the system clock pulse (C) ,
and the timing pulse (Bo) are the clock -generator outputs.
The preclock pulse (G) is derived from the leading edge of the Ts
TIN
TS
(Normal)
J
TS
(Fast Search)
BG
Fig. 14 -19. Timing diagram of fast- search logic.
waveform and is coincident with this leading edge. Logic circuits are
preset by the G- pulse, which occurs prior to system clock pulse C and
therefore assures that switching transients subside before system switch-
ing takes place. A further consideration for this arrangement is the fact
that the logic components used in the circuitry possess inherent and finite
transfer delays, which could prevent the correct completion of switching
sequences during the presence of a system clock pulse.
The Ts -pulse is applied to an inverter. After differentiation, the leading
edge of the Ts waveform produces a negative -going pulse, which is used
to trigger a 20- microsecond one -shot multivibrator. The inverted output
represents pulse G and is routed to the slow- motion quantizer and the
slow- motion logic.
VIDEO DISC RECORDING SYSTEMS
I nputs
Fs I
TS
Output s
Ts
6G
Fig. 14 -20. Waveforms of clock -pulse generator.
The system clock pulse (C) is derived from the trailing edge of the
Ts- pulse. Its timing is such that the leading edge of the C -pulse is coinci-
dent with the trailing edge of the Ts- pulse. The timing of the C -pulse
therefore is such that it occurs approximately 600 microseconds ( the dura-
tion of the Ts- pulse) after the G- pulse. The C- pulses are used to time the
signal- switching and carriage -drive pulses.
The BG -pulse is a divided -by -two version of preclock pulse G and is
phased by field- identification pulse FS. Phasing of BG by means of Fs
causes even fields to be recorded by heads A and C and odd fields to be
recorded by heads B and D. This condition is valid for the normal record
mode only, since the alternate -field record mode does not allow recording
of even fields. The BG waveform is a binary 1 during even fields and a
binary 0 during odd fields.
Slow -Motion Quantizer -(See timing diagram, Fig. 14 -21.) The slow -
motion quantizer takes slow- motion reference waveform A (generated in
the control unit and routed through the alternate-field logic as As) and
converts it to ZG. Output ZG is a rectangular wave, each zero crossing of
which is coincident with a G- pulse. The number of zero crossings of ZG
I n put
11111111111111111111111111111111
u ;tput
Nor
SiOw 'aOtÍP.
Fig. 14 -21. Waveforms of slow- motion quantizer.

motion by circuits on the carriage logic board and re -enter the switching
logic board as waveforms EAK and ECK.
Head Retiming Logic -(See timing diagram, Fig. 14 -24.) The head
retiming logic reclocks all the head pulses against waveform C prior to
the application of these pulses to the head switcher. The four head -pulse
inputs and their complements are applied to preset inputs of JK flip-flops.
The complements of the input pulses are obtained from inverters made up
of NAND gates. All four flip-flops are clocked by waveform C and there-
fore produce head -pulse output changes coincident with the leading edge
of clock pulse C. The resulting output pulses-EAC, EBC, Ecc, and Enc-
ase routed through the harnesses to the rf switcher and switching logic
board in the dsu. Pretimed waveform EDG is also passed through an in-
verter. Thus EDG is routed to the control unit for synchronization of the
indicator clock.
Half-Line Delay Logic- During playback in any mode but normal-
speed forward, an odd field may be gated through the reproduce electronics
at the time an even field is required; or, conversely, an even field may be
gated on when an odd field is required. Since this would destroy the pic-
ture interlace, it is necessary to convert odd fields to even fields, and vice
versa. Conversion of one to the other is accomplished by a half-line delay
of the video signal during the horizontal- scanning interval (but not the
vertical- synchronizing interval) of each field requiring correction.
The half-line delay logic generates waveform R', which controls the
half -line delay. The delay is switched out when R' = 1 and switched in
when R' = 0. To do this, the logic determines (A) whether an odd or even
field is required and (B) whether an odd or even field is being reproduced.
If a correction is required, R' becomes 0, and for no correction R' is 1.
Also, during the vertical- synchronizing internal, R' always is held to the
value 1. Waveform BG determines what type of field is required, since BG
is backed up to reference sync and always has the value 1 during
even
fields and the value 0 during odd fields.
J
The DG pulse determines what type of field is actually being reproduced.
In normal speed, DG = BG, so odd fields are always reproduced when odd
4v
(
OV- -L
EBOI
(
V
ov I I
}
4v-
OV
Note. Waveforms not to scale.

Fig. 14 -24. Timing diagram of head- retiming logic.
fields are required, and even fields are always reproduced when even fields
are required. The half-line delay, therefore, remains switched out. In slow
motion, odd fields are reproduced when DG = 0, and even fields are re-
produced when DG = 1.
Waveforms BG and DG are connected to a NAND gate, and BG and D,
are connected to another NAND gate. The outputs are tied together and
form an exclusive OR function labelled R'. This function follows the
Boolean equation:
R'=BG DG+DG' BG
or,
R' = BG DG + BG DG
Thus, R' equals binary 1 if the half-line delay is not required.

Alternate -Field Logic and Switch SI -The alternate -field logic circuit is
located on the fast - search logic module in the esu. In the alternate-field
mode, only one field of each frame is recorded, providing twice the normal
recording time. In playback, since half the fields are absent, motion appears
to be at double the rate it would be for normal recordings.
The alternate -field switch (S1) controls three functions:
(A) Provides Ale, the alternate -field -mode command, to the control
unit. The level of AF- is logic 1 in the alternate -field mode and
logic 0 in the normal mode.
(B) Provides the K' waveform to the chroma -inverter logic. The K'
signal is K (the quantized forward /reverse command) in the
normal mode and is 0 in the alternate-field mode.
(C) Provides the Ra waveform for the half -line delay and chroma -
inverter logic. In the normal mode:
Ra =R' =BG DG + B0 D0
In the alternate -field mode:
R =BG
Chroma- Inverter Logic -The chroma - inverter logic circuitry is on the

chroma -inverter module, which is located in the esu. The chroma inverter
controls the chroma phase continuity of the reproduced signal, which
would otherwise be interrupted in any mode other than normal-speed for-
ward playback. The chroma -inverter logic controls the action of the chroma
phase inverter by means of a binary waveform which changes level each
time the chroma phase is to be reversed.
For recordings made in the normal record mode, chroma phase must be
reversed each time the half -line delay is inserted in forward -motion play-
back. In reverse -motion playback, the chroma phase is also reversed each
time the heads step.
In the alternate -field mode, the action of the chroma inverter is the
same for forward and reverse motion. The chroma phase must be reversed
each time the half -line delay is inserted and each time the heads step, unless
the two occur at the same time. The chroma- inverter control waveform is
designated Cg. The chroma- inverter logic has been explained previously.
Disc Servo Unit Logic
The disc servo unit logic is located on the carriage logic Al and car-
riage logic A2 modules, which are located in the dsu (Fig. 14 -25) . The
purpose of the logic is to provide the head -stepping command pulses
(F- pulses) to the carriage -motor drive amplifiers. This logic maintains
proper head synchronization for all motion speeds, forward and reverse
directions, and inward and outward travel of the heads across the disc. The
arrival of the head carriage at the outermost or innermost tracks of the
discs is detected by means of photocell end -stop switches, which control
the reversal of head-carriage motion at these points.
Electronics Unit
EAG EBG ECG DG
lEAK G
ECK 0 IC
EAG
EBG
Reverse -Motion
ECC Carriage
Logic EEG Logic
IA21
FAG IA21
K EAKFCK ECG
IEBG FBG FOG
Carriage Retiming
Carriage Control F0
Logic
Logic F
IA11
(Al) c
Fd
Fa Fbc Fcc Fdc
A l A
Jc X N
EBG
r M
X0 YdX
Xpa Xc Xd Ya Yb Yc
X X0
Oco
XC Xc Carriage
Carriage
Error ü0 Carriage
Reversing
Logic XD End Stops Xd Correction rd; MDA
YA Ya
IA21 Logic rTa
YB Yb (All rt-ci
Yc
YE
YD Yd
Fdcì
Fig. 14 -25. Logic block diagram of disc servo unit.

, one head is recording, one

see Table 14 -6) is
In the record mode (
erasing (in preparation for recording the next field) , and two are step-
ping to the adjacent track during each field interval. For example, during
field 1 ( when EA = 1) , head A is recording, head B is erasing, and heads
C and D are stepping. Similarly, during field 2, E1 causes head B to record,
head C to erase, and heads D and A to step. Then during each cycle of
four fields, each head erases, records its field, and steps two tracks before
repeating the cycle. These pairs of stepping pulses, the F pulses, are gen-
erated in the block labelled "carriage logic" in Fig. 14 -25.
Table 14 -6. Head Sequencing, Normal Record
Field Field Field Field

1, 5, 9, 2, 6, 10, 3, 7, 11, 4, 8, 12,
etc. etc. etc. etc.
Head A Record Move Move Erase
Head B Erase Record Move Move
Head C Move Erase Record Move

Head D Move Move Erase Record
The motion of the heads across the discs is shown in Fig. 14 -26 for
normal -speed playback. The fields are displayed horizontally, the track
number vertically. Head motion is identical to that described above for
the record mode. During field 1, head A is reproducing on track 6. During
field 2, head B reproduces on its track 6, while head A moves one track to
track 5. During field 3, head C reproduces its track 6, head B moves to
track 5, and head A moves to track 4. The heads move toward track 1 in
this manner, each reproducing a field in turn and then moving twice to
the next even-numbered track. After reproducing field 9, head A steps
once, then reaches its end stop. Further motion of head A in that direction
is inhibited by the carriage error -correction logic. By the end of field 13,
all heads have reached their end stops at track 1. At this point, the direc-
tion of head -carriage motion is reversed, and head A begins moving away
from the end stop, this time reproducing the odd -numbered tracks, fol-
lowed by heads B, C, and D in the manner described above. When the
other set of end stops is reached, a similar carriage reversal takes place,
and the heads once again reproduce the even -numbered tracks.
If, instead of forward motion, the operator selects reverse -motion play-
back, the recorded tracks must be played back in the sequence opposite to
that in which they were recorded. Such a reversal of motion is shown at
the end of field 21 (Fig. 14 -26) The direction of carriage motion is re-
.
versed at this point, and the E pulses are made to reverse their sequence
(Ec- EB- EA -ED, instead of EA- EB- Ec -ED). This reversal of E -pulse sequence
is accomplished by interchanging the EA and Ec pulses. Upon selection of
Fig. 14 -26. Head -stepping diagram.

reverse motion at field 21, field 22 is reproduced by head B, field 23 by

head A, field 24 by head D, and so on.
In reverse -motion playback, carriage reversal at the end stops is accom-
plished in a manner somewhat different than for forward motion. Follow-
ing reproduction of track 24, head D steps twice and arrives at its end
stop, followed in turn by heads C, B, and A. When all heads have ar-
rived, the direction of carriage motion is reversed as before. Unlike the
case for motion in the forward direction, each head is now allowed to
take just one step away from the end stops before reproducing a track,
thereby maintaining a mirror image of the forward- motion carriage -
reversal sequence. To summarize, the carriages, in forward motion, make
single steps when arriving at the end stops, and take two steps on leaving.
In reverse motion, they take two steps when arriving at the end stops,
and one step when leaving.
At the end stops, the carriages are automatically resynchronized with
one another to correct for any possible carriage stepping errors. If, for
example, a head has gotten out of step due to a momentary power failure,
the inhibiting action of the end -stop photocells holds all heads at the end
track until the last head has arrived. The logic then ensures that the heads
leave in correct order, A -B -C -D in forward motion and D -C -B -A in re-
verse- motion playback.
In slow-motion playback, head motion follows the sequence described
above, except that each time a field is repeated, no carriage- stepping pulses
are generated.
Carriage Logic A2 Module -The carriage logic A2 module in the dsu
contains: (1) the carriage logic, (2) the carriage reversing logic, and
(3) the reverse motion logic. It should be clearly understood that reversal
of carriage motion at the inner and outer track limits is not a reversal of
video motion, as the reproduced video still progresses in a forward video
motion. Forward video motion is identified by a binary -1 level of signal K
and by the head and carriage sequence A- B -C -D. Reverse video motion is
characterized by a binary -0 level of signal K and by the head and carriage
sequence C- B -A -D. The A and C functions are interchanged for the
reverse modes to produce reverse motion while maintaining correct odd -
to -even field relationships.
The carriage logic receives four inputs. These inputs are head pulses
from the switching logic board in the esu and are labelled EBG, EEG, EnG,
and EAG. These signals are gated in pairs by the use of NAND gates to
produce carriage- motion signals FAG, FRG, FCG, and FDG. Signals FRG and
FDG are routed to the output. Signals FAG and Fo0 are routed into the
reverse -motion logic. The Boolean equations for the four signals are:
FAG = EsG + ECG
FBG = ECG + EnG
FcG = EDG + EAG
FDG = EAG + ERG
It can be seen from these equations that each carriage will step twice
during each series of four E pulses, satisfying the stepping requirements
for forward -motion operation. Further processing of these signals is re-
quired to produce carriage- stepping pulses which satisfy the requirements
for carriage movement during reverse motion and while negotiating the
end stops.
Carriage -Reversing Logic-The carriage- reversing logic generates the
carriage- reversing command (M) which governs the direction of motion
of the carriages. Carriage motion may be inward or outward for forward
or reverse video motion. Inward motion is defined as carriage travel from
the periphery toward the hub of the discs; outward motion is the oppo-
site. The arrival of a head at the outer track is detected by photocell X (at
the inner track by photocell Y) for each disc surface. The carriage-
reversing command (M) is a function of two signal sources, the motion -
reverse command (K) and the end-stop switch signals.
The reverse- motion logic circuitry generates quantized reverse -motion
command K and produces the reversal of the channel switching and
stepping sequence (A-B -C -D changed to C- B -A -D) during reverse mo-
tion. This is accomplished when K = 1 by interchanging the EA pulse
with the Ec pulse, and the FA pulse with the Fc pulse.
Reverse- Motion Command K- Command K is equal to the P2s com-
mand quantized by head signal EBG and preclock pulse G. Command K is
also inhibited from changing while the carriages are at their end-stop
switches, as motion-reverse switching coincident with carriage end -stop
reverse could cause errors in the carriage reversing process.
Head Switching Sequence in Reverse Motion -Reverse video motion
is obtained by reproducing the video in a sequence opposite to the
record
sequence. Substitution of the channel -A timing signals for the channel -C
signals and vice versa converts the sequence from A -B -C -D to C -B -A -D
for reverse video motion. The head pulses for channels B and D are not
affected by reverse motion, but the pulses for channels A and C are.
Carriage Stepping Sequence in Reverse Motion -The reasons for ex-
changing the channel -A and channel -C pulses described in the preceding
paragraph for the head logic apply also for the carriage pulses. Output
signals FAK and F(.K are derived in a manner similar to that used for the
head logic, except that carriage- motion pulses FAG and FVG are used in-
stead of head pulses EAG and ECG. Four gates are contained in assembly
A2, which is used in this circuit. Pairs of gates have their outputs tied
together, performing the function of an OR gate. The Boolean equations
for each of the output signals are as follows:
FAK = (FAG ' K) +(FcG'K)
FUK= (FcGK) +( FAG g)
where,
FAG = Esc + ECG
and
Fco = EDG + EAG
Carriage Logic Al Module-The carriage logic Al module of the dsu

contains: (1) the carriage control logic, (2) the carriage retiming logic,
and (3) the carriage error-correction logic. Outputs from this module are
control pulses representing inward and outward carriage- motion pulses.
The carriage- control logic controls the movement of the carriages away
from the end stops in such a way that they start in the sequence A -B -C -D
(with A first) for forward motion, and in the sequence D -C -B -A (with D
first) for reverse motion. In addition, for reverse motion, each carriage is
allowed to move only one track, rather than the normal two, as it leaves
the end stops. This control is obtained by gating off the F pulses to any
head whose motion must be inhibited.
See Fig. 14 -27. When all heads are at their end stops, the direction of
the carriage motor is reversed by the carriage -control logic, and head A
then steps away from its end stop. Head B, which otherwise moves along
with head A, must be inhibited in order to preserve the A -B -C -D sequence
of head motion.
Carriage Motion Reversal
Fe i
F D'
E I
Time
Notes: E End Stop Activated
I Stepping Inhibited
Fig. 14 -27. Carriage pulse sequence, forward video motion.
See Fig. 14 -28. When all heads reach their end stops and the direction
of carriage motion is reversed, each head, in the sequence D- C -B -A, must
take only one step away from the end stops before resuming normal 2 -track
stepping. Head D, the first to leave, automatically moves only one track
because carriage reversal takes place after what otherwise would have
been its first step.
Carriage Retiming Logic -The carriage retiming logic changes the
timing of the carriage signals from preclock time G to system clock time
C, then samples them with the J pulse to produce the short -duration car-
Carriage Motion Reversal
V
-
FA
Time
Notes. E ° End Stop Activated
I = Stepping Inhibited
Fig. 14 -28. Carriage pulse sequence, reverse video motion.
Normal Motion Slow Motion
EAK
FBG
ECK
EDG
k 11111111111111 I I I I I I I
FA'
FAG' I I I I I I I I I I
F1'
F BC' I I I I I I I I I I I
FC'
FCC' I I I I I I I I I
FDC' I I I I I I
Time
Fig. 14 -29. Head and carriage timing waveforms of carriage -retiming logic.
riage pulses FA(', FRC', FCC', and Fix,' It employs binary devices and asso-
ciated NAND gates and inverters. The four channels operate in a similar
manner. See Fig. 14 -29 for waveforms of the retiming logic.
Carriage Error- Correction Logic -The carriage error- correction logic
serves to inhibit the further motion of each carriage as it reaches an end
stop. When all four heads reach their respective end stops and the direc-
tion of carriage motion is reversed by the carriage reversing logic, the
carriages are allowed to step away from their end stops in correct sequence
as determined by the carriage control logic. Output pulses Fac,,, Fb,.,,, Fee.,
and Fac,, control the outward movement of the carriages. Pulses Fad, Fbei,
Frei, and Fdei control the inward movement of the carriages.
14 -6. THE AMPEX HS -200 TELEPRODUCTION SYSTEM
The Ampex HS -200 teleproduction system is an instant- replay disc re-

corder that can be operated manually or automatically to provide a variety
of special effects, editing, and animation possibilities. This system offers
the same freedom to create the special effects usually limited to film, but
at lower cost and greater speed.
The HS -200 system is a natural outgrowth of the color disc technology
of the HS -100B instant -replay recorder /reproducer. Like its predecessor,
the HS -200 reproduces material at normal, fast, and slow speeds -even
to stop action -in both forward and reverse, with instant replay of key
action. Versions are available which are fully compatible with monochrome
or NTSC and PAL color video signals.
NOTE: The HS -100B is the video disc system used in the HS -200 tele-
production system.
Features
Because of the instant review feature of the HS -200 teleproduction sys-
tem, animations or other edited program material are available for review
in full color immediately after recording. When the material is reviewed
in its finished form and corrections are necessary, all or any part of the
material (even to a single frame) may be replaced. The preprogramming
feature of the HS -200 system permits automatic editing of complete pro-
grams, such as commercials and animation sequences.
System Console -The console contains all the primary operating con-
trols for the system. The primary controls and indicators are arranged by
function on six panels: record panel, memory panel, slow- motion panel,
dissolver panel, numerical -display panel, and patch panel. In addition, a
picture- monitor panel and a waveform -monitor panel are provided.
Variable Frame Increment (Animaation) -The HS-100 has a total ca-
pacity of 900 frames, which may be recorded in one operation or in a
selected number of frame increments between one and 900. By this means,
it is possible to complete in a relatively short time an animation sequence

that would ordinarily require hours of cutting, reviewing, and editing.
Animation can be preprogrammed and requires only two passes of
video tape on the video recorder. On the first tape pass, the vtr records the
cues at appropriate places on the tape. The HS -200 is used on the second
pass to compile the animation sequence automatically. In the second pass,
each vtr cue commands the HS -200 to record the preset number of frames.
This second tape pass has been simplified to permit a technician to com-
plete the assignment without the assistance of the editor. With all ma-
chines operating in any one recording session, correct color editing is
always maintained between one camera and the HS-200 or between one
vtr and the HS -200, provided that neither the HS-200 nor the vtr trans-
port is turned off. Unless the vtr is provided with 15 -Hz editing pulses,
there will be only a 50 -50 chance of correct color editing each time the
transport is turned off.
Time Lapse -Time lapse is an extension of the animation feature.
The HS-200 records a predetermined number of frames; then a reason-
able number of frames is allowed to pass unrecorded. Replaying this se-
quence at normal speed will produce the effect of time lapse. An auto-
matic time -lapse unit is available as an accessory.
Random Access to Any Frame -The system includes a numerical -entry
keyboard and a numerical- readout frame counter which provides every
frame with an identity. Any number from I to 900 can be called up,
stored, and then displayed. When using the automatic search feature, the
HS -200 will retrieve the desired frame and place it in a freeze condition.
This frame may be used as an individual frame or the first frame of a
sequence.
Freeze Frame or Field -On the HS -200, a freeze picture can be ob-
tained in two ways. One method is by repeating one recorded field; the
video is fed through a half -line delay during every other field interval to
produce a frame. Because each frame is made of two identical fields, there
is a loss of definition. The other method available on the HS -200 is by
repeating two adjacently recorded fields to make up the frame. In this
way, full definition is achieved. These two modes are called the field and
frame modes, respectively.
If a freeze picture of some recorded fast action is required, the field
mode is normally used because the frame mode would produce a blurred
picture due to the fact that the frame was made of two pictures recorded
1/60 second apart. The frame mode is used during animation to provide
full definition.
Slow motion is produced by repeating fields or frames. The same choice
of repeating frames or fields is available in slow motion.
Time -Controlled Freeze Frame-Any frame can be accessed and held in
freeze for any time period between 2 and 900 frames before another
mode, such as switch or forward or reverse motion, is actuated.
Automated Dissolves of Variable Length -The HS -200 contains a digi-

tal dissolver. The dissolve rate can be preset to any one of the following:
cut, 4, 8, 11, 16, 22, 32, 43, 64, 86, 128, and 256 frames. A dissolve be-
tween any two signal sources ( one of which would normally be the out-
put of the HS-200) will take place at the push of a button, at the receipt
of a remote cue, or at any frame selected on the HS -200.
Matting and Keying -A matting and keying amplifier, which comes as
an accessory, allows adjustment of luminance, chrominance, and saturation
of the matting signal. This device will key or mat at the push of a button,
at the receipt of a remote cue, or at any selected frame of the HS-200.
The HS -200 enables the user to animate any keying pattern and record it
on tape for future use.
Sequential Programming Control Circuits-The patch -panel unit built
into the HS -200 system can receive up to eight switching, mixing, and
editing commands. These commands are stored until cued. Once the pre-
programming is prepared, the HS-200 will assemble an entire segment
automatically.
Variable -Speed Playback -As with the HS -100, all recorded material
can be played back at any slow-motion speed between freeze and normal
in either the forward or reverse direction. There are also preselected slow -
motion speeds of exactly I/2 and 1/4. In addition, the HS -200 allows these
features to be programmed through the patch panel. With the alternate-
field mode, alternate-frame mode, or time -lapse unit, motion can be
speeded up above normal speed.
Two -Way Remote Control and Rehearsal-The HS -200 can receive
cues from another HS -200, or from a vtr, in the form of logic commands
or tone bursts, and act upon them through the patch panel. It can also
send signals to another HS -200 or vtr. This, for example, permits prepro-
gramming of an A-B roll consisting of cuts, dissolves, and slow -motion
effects between two HS -200's or one HS -200 and a vtr.
Built -In Monitoring -The system console includes a waveform moni-
tor and a video switch to permit the following points of the system to be
monitored: input video, Colortec output, disc output, dissolver output,
key -amplifier output, system output, and assembly vtr output; in addition,
there is provision for one spare.
NOTE: The disc servo unit, electronics signal unit, and output processing
unit are referred to collectively in this text as the HS -100 or HS -100B
slow -motion video recorder /reproducer. Also included as part of an HS-
100 or HS -100B is a small control unit, previously described. If the con-
trol unit is part of an HS -100B, it can be replaced by the HS -200 console
to form the HS -200 system.
Assembly -The simplest concept of motion -picture editing (either film

or tape) is that of assembly, the putting together, in sequence, of various
scenes available from prerecorded material, or the production of scenes
in the required order at the time of editing. The process of assembly from
prerecorded material involves locating the desired sequence and attaching
it to the preceding one either physically, photographically, or electroni-
cally. This is actually the only way the operation can be done with film
with any degree of precision or artistic flexibility. With either tape or disc,
however, assembly and recording can be accomplished concurrently by
producing the scenes in order and recording each one to its desired tim-
ing and making the desired transitions (Chapter 10) The HS -200 adds
.
an order of magnitude to the ease, speed, and precision with which this
can be done, and progress may be reviewed as the operation is carried out.
Insertion-Insertion of new material into previously recorded material
is a common requirement. Accurate placement of the insertion with regard
to preceding and following material requires the ability to determine the
"in" and "out" edit points of both the new material and the material into
which it is to be inserted (Chapter 10) The HS -200 offers two possibili-
.
ties for carrying out such an operation: (1) The material to be inserted is
recorded on the HS -200 and cued accurately to the desired "in" point and
then electronically edited into the prerecorded sequence on tape from
cues previously placed and precisely adjusted. (2) The original material
is recorded on the HS -200, and the new material is inserted at the desired
points from any other source, once again with one-frame accuracy of
placement.
Cutting -The cut (an instantaneous change of scene) is the simplest
of video transitions both physically and subjectively; timing is the only
variable. With film, proper timing is accomplished by selecting the frame
at which the cut is desired. The HS -200 offers this facility by allowing
examination of material at a frame -by -frame rate, and it also allows realis-
tic preview of the cut before it is accomplished.
Dissolving -The dissolve is a more complex transition, requiring the
fading out of the first picture simultaneously with the fading in of the
second one. Timing again is the critical point, with respect not only to
when the dissolve occurs, but also to how long it will last. With film, this
is an operation that is done in the optical printer with precise control of
the number of frames involved, and therefore must be entrusted to the
lab technician. With the HS -200 and an accompanying video source, dis-
solves may be achieved with the same accuracy and automatically con-
trolled as to the length. Again, the major advantage offered by this system
is the ability to preview or rehearse the effect with the possibility of
making desired adjustments before actually performing the final transition.
Fading -The fade involves simply fading from picture to black or
vice versa. It is easily accomplished in a number of ways. It can be done
at the time of recording or during the editing process. The advantage
offered by the HS -200 in this operation is accuracy of timing.
A -B Roll -The A -B roll technique, commonly used in film production,
involves dividing the production into "odd" and "even" scenes. The odd
scenes are placed in sequence on one video source and the even ones on
another. With proper timing between the two sources, odd and even
scenes are then combined into the desired final form. With the precision
of control available, this is easily accomplished with two HS-200's or by
one HS-200 and a second video source.
Keying -In this process, one picture signal is inserted into another, e.g.,
a title into a background scene. In film, this is a laborious process requir-
ing time and expense. In television, it is achieved electronically by com-
bining two video signals. The main use of the HS -200 in this process is as
the source of the signal to be keyed. Because of the frame -at -a -time play-
back capability of the HS -200, material to be keyed may be recorded on
succeeding frames and keyed into the background as desired and changed
on cue. It is possible to record animated titles or products on the system
and key them into other material with full control of movement and
timing.
Matting -This process is the same as keying as far as the HS-200 is
concerned.
Superimposure- Superimposure is a process that combines video ma-
terial from two sources. The end result of superimposure is similar to that
of keying or matting, although the effect is not as crisp. The HS -200 acts
as one of the sources, probably for printed material.
Time Lapse-This technique, which might also be called "time com-
pression,' allows the reproduction of lengthy events to be condensed as
much as desired. The blooming of a flower, for instance, can be repro-
duced in a period of five seconds. This effect is accomplished by taking
picture frames at selected intervals during the event being recorded, rather
than recording it continuously. When the reproduction is played back con-
tinuously, time compression is achieved. In other words, if one frame is
recorded every five seconds during a 15- minute process, the final record-
ing will have a total of 180 frames. If these frames are played back at the
normal rate (30 frames per second for television or 24 frames per second
for film), the whole event can be viewed in six seconds. This is done on
film by exposing one frame (either manually or by a timer) at every
selected interval. The same technique is possible on magnetic recordings
with the HS-200. Two advantages accrue: (1) the reproduction is imme-
diately reviewable, and (2) playback is not locked to any particular frame
rate (as it is in film) , allowing the length of the event to be manipulated
to fit any time segment.
Simulated Time Lapse-This interesting technique, which produces the
effect of time lapse, is achieved by combining the backward- motion and
single -frame recording capabilities of the HS -200. For example, a potted
plant may be placed on camera and a series of frames recorded with a
little of the plant snipped off between each "take" until the plant has been
cut entirely away. If this sequence is played in reverse, the plant seems to
grow before the camera. This technique can be applied to a number of
given instant. In freeze operation, for example, the frame counter is sta-
tionary. When the HS -200 is either turned off or placed in the standby
mode, the system may develop a counting error. To resynchronize the
frame counter after turn -on, the controls are set for normal forward op-
eration, and all four heads are allowed to move to the outer edges of the
disc.
There are eight stores accessible through the memory panel. The num-
ber in the store being accessed at any time is shown on the store display.
However, the particular store being accessed on the memory panel is indi-
cated by an illuminated store push button.
MODE
STANDBY RECORD PLAYBACK
ALT FRAME FIELD
OPERATE
FRAME FIELD
POWER FRAME FRAME

TIMER
ALT FIELD FRAME STORE
HOLD
SYNC
LIGHT
O
RECORD RECORD START STOP
TIMER
READY
R T S S
STANDBY/OPERATE Controls power to disc motor and

hour meter
MODE Selects recording /reproducing mode
POWER ON /OFF Controls system per

TIMER STORE HOLD Stores number of frames displayed
in Store Display for automatic timing
of recordings
RECORD READY Switches system out of reproduce

mode and prepares it for recording
RECORD TIMER Selects manual or automatic timing

of recorded intervals
START Initiates recording sequence

STOP Terminates recording sequence
Fig. 14 -30. Record panel on HS -200 control console.
Record panel- Operations performed on the record panel (Fig. 14 -30)

control the major recording functions of the system.
The on position of the POWER ON /OFF switch supplies all required
power to the system; however, even in the off position a small low- current
28 -volt relay supply remains on. To remove all power, the main breakers in
the HS -100 (CB -1, rear connector panel of esu) and HS -200 console (in-
side lower left front panel) must be switched off.
The OPERATE /STANDBY push button controls the power to only the disc
motor and hour meter, and has no effect on the remaining circuits, cotn-
puter logic, or information in stores. Setting this control to Standby will
save unnecessary wear on the heads and discs.
The green SYNC indicator lamp is in parallel with the lamp of the disc
servo unit. When the lamp is fully on (not flickering), in indicates to the
operator that the disc servo is in sync. Neither a correct recording nor a
correct reproduction can be achieved unless this lamp is on.
The RECORD READY control takes the system out of the reproduce mode
and prepares it for instant recording in response to a record command.
When the RECORD READY push button is pressed, the SYNC lamp will flash
momentarily, then glow steadily, indicating that the system is ready to
record. NOTE: The system will not switch to record ready from the re-
verse reproduce condition; it must first be placed in forward.
The start, stop, and record timer functions are best explained by de-
scribing manual, automatic, and programmed -sequence recording opera-
tions. To make a manual recording. the record timer must be off and the
system must be in record ready. With the green SYNC light steady and the
START push button pressed, the system will begin recording on the next
15 -Hz color editing pulse. The system will stop recording when the STOP
button is pressed. To make each successive recording, press the START
button and then the STOP button.
To record a preset number of frames automatically, the record timer
must be on and the system must be in record ready. When the START
button is pressed, with TIMER STORE HOLD off and SEQUENCE STORES
(Fig. 14 -31) off, the HS-200 will wait for the next 15 -Hz color editing
pulse, after which it will record for the number of frames displayed on
the store display. The same number of frames is recorded each time the
START button is pressed.
Up to eight different recording lengths can be programmed in sequence
instead of one at a time. The length of the first recording in frames is put
in store 1. The length of the second recording in frames is put in store 2,
and so on up to store 8. When this is completed, the RECORD TIMER, RE-
CORD READY, and SEQUENCE STORES push buttons are all set to the on
position, and the TIMER STORE HOLD button is off.
The first time the START button is pressed, it will cause the number of
frames in store 1 to be recorded. The second time it is pressed, it will
cause the number of frames in store 2 to be recorded, and so on. As al-
ways, the manual operation of the START push button can be replaced by
a cue from a vtr (by feeding a tone -burst cue, for example, and patching
from Remote VTR A to Start Record) .
Pressing the TIMER STORE HOLD push button to the on position stores
the number shown on the store display in a ninth store called the timer
store. The number will be stored there until the TIMER STORE HOLD button
is pressed to off. If the system is in automatic record, it will record for
the number of frames in this ninth store. With a pencil, the operator
should write the number stored in the space provided above the TIMER
STORE HOLD push button. After the number is stored, the display may be
changed without affecting the number stores. The use of the timer store
hold feature is recommended, instead of methods described which do not
use this feature, for making an automatic recording. If the wrong store
is accessed by mistake when the TIMER STORE HOLD push button is off,
2 3 4 5 6 7
STORE STORE STORE STORE STORE STORE STORE STORE
ENTER
ENTER
CUE
C E
17 8 9
5 6
2 3 SEQUENCE
.11
STORES
o 0
R
SEARCH
START
PROGRAM
S P
STORING FUNCTION
STORE 1 -8 Store Selector
ENTER CUE Enters displayed frame counter into

store
ENTER Enters keyboard selection into store
Keyboard Selects three -digit number for stores
OPERATIONAL FUNCTIONS
SEQUENCE STORES Returns program selected on patch

panel to beginning
SEARCH Initiates search and display operations
for the frame number shown in Store
Display
START PROGRAM Begins program selected on patch panel
Fig. 14 -31. Memory panel on HS -200 control console.
a recording of incorrect length may be made. This is impossible when the

timer store hold feature is used.
The MODE selector switch selects the recording and reproducing modes
of the HS -200. The four switch positions are labelled to describe the mode
of recording:
1. ALTernate FIELD
FIELD Reproduce
2. FRAME
3. FRAME
4. ALTernate FRAME
FRAME Reproduce
Recordings made in positions 1 and 4 must be reproduced in positions 1

and 4, respectively. Recordings made in positions 2 and 3 may be repro-
duced in either position 2 or 3.
When recordings are made in position 2 or 3, every field and frame is
recorded on the discs. These recordings can then be reproduced in two
ways. In position 2, reproduction is in the field mode, in which slow -
motion or freeze pictures are produced by repeating individual fields and
using a half -line delay to obtain the interlaced frames. This provides fox
best picture resolution when there is motion in the picture between one
field and the next, as in a fast -action scene. In position 3, reproduction is
in the frame mode, in which slow- motion and freeze pictures are produced
by repeating complete frames. This provides for best picture resolution
when there is no scene motion between fields of a given frame, as in an
animated scene. The frame mode gives full vertical resolution, but the field
mode does not.
In position 1, the HS -200 records every other field (one field of each
frame) Playback is in the field reproduce mode, using the half-line delay.
.
This produces some loss of vertical resolution in the picture.

In position 4, the HS -200 records every other frame. Playback is in the
frame reproduce mode without using the half-line delay. This produces
some loss of motion resolution but no loss of vertical resolution in the
picture.
Memory Panel -The push- button controls on the memory panel (Fig.
14 -31) are divided into two groups. The first group-STORE 1 through 8,
keyboard, ENTER, and ENTER CUE -is used to enter and store frame num-
bers. In the second group, SEQUENCE STORES changes the operational mode
of the eight store buttons, and SEARCH and START PROGRAM are operating
controls.
System Operation
We will go through a brief description of operating conditions and the
philosophy associated with system control. Detailed step-by -step procedures
are provided in the instruction book for the HS -200 system. Obviously,
space does not permit such details in this text, and only fundamentals im-
portant to a general understanding of operations procedures are presented.
Assemble- Because of the precision with which material can be manipu-

lated on the HS -200, it is not necessary to assemble in sequence once the
accuracy of segments and overall timing has been determined. For exam-
ple (refer to Fig. 14 -32) If it is known that the third scene in a sequence
:
will occur from the 18th to the 24th second of the total time, the scene
can be programmed on the HS-200 to start recording at frame number 560
and stop 180 frames (6 seconds) later, at frame number 740. If it is being
recorded "live," the action is cued, and the HS-200 is started manually.
Only the number of frames desired will be recorded, and the next scene
can then be recorded in a similar manner by again adjusting the number of
frames to be recorded. If the material is coming from a video -tape re-
corder, a cue may be placed at the "in" point on the tape, and this will start
the HS -200 in record automatically. In this way, precise timing adjust-
ments of scenes and cuts are possible.
0 Sec 10 Sec 18 Sec 24 Sec 30 Sec
Scene I Scene II Scene III Scene IV
Frame 20 Frame 320 Frame 560 Frame 740 Frame 20
IIIIIIIIIIIIIIIIIIIIIIIIIIIIIII
300 Frames - -f- 240 Frames 180 Frames -1 -180 Frames
Total Sequence
Fig. 14 -32. Assembly of scenes.
Now, if it is desired to go back and record scene II, just determine the
length of scene II (in frames) subtract that from the starting point of
,
scene III, use the resulting frame number as the starting point of scene II,
and program the HS -200 to record the number of frames in scene II.
An entire sequence would be recorded as follows:
A. Cue the HS -200 when the frame counter displays 560; then program
the system to record 180 frames. This will record scene III.
B. Without changing the frame number on the HS -200 or the number
of frames to record, record scene IV, which is exactly the same length
as scene III and follows directly after it.
C. Next, to record scene II, which is 8 seconds (240 frames) in length,
cue the HS -200 at frame count 320, which is exactly 240 frames
before the beginning of scene III. Set the record timer for 240
frames and record.
D. Finally, record scene I in an identical manner, cuing the HS -200 at
frame readout 20 and setting the record timer for 300 frames (10
seconds) , the desired length of scene I.
Frame 20 was arbitrarily chosen as 20. Because the HS-200 records in a

complete loop, the frame count in scene IV goes to a count of 900 followed
by 001, etc.
Insert- Inserting new material into previously recorded material may
be done in the following ways.
A. Inserting into material already recorded on the HS-200. This is
similar to the assemble process explained above.
B. Inserting from an HS -200 onto a vtr
Record the original sequence on an Editec -equipped vtr. Place Editec

cues at the desired "in" and 'but" points on the tape. Record the material
FROM T TO
Remote VTR @A © B ©C ©A OB ©C Remote VTR

Remote Disc ©A Q0 B OO C ©A OB OC Remote Disc
Remote Aux ©A OB OC @A ED ©C Remote Aux
Start Program © Start Program Oo Record Ready

Sequence Stores © Qo Sequence Stores Oo Start Record
© Stop Record
Timer Out A O 0 A Timer In © Search
Search © Timer Out B © © B Timer In
Sequential Disc OR GATES

Normal Inputs
1
Address Oo Out
Address
©1 1O © Slow Motion 1 O O 0 o
2O ©
OO 2
Oo 3 3 Qo ©
Slow Motion
Slow Motion
2
3
00 o o
00 4 4 O Qo Freeze 00 o o
O 5 5O O Single Frame
O o o
©6 6 0 ® Forward
07 70 Oo Reverse O cap®
BO
OB
00 o o
Seq Patch
Address Panel
O O Oo Dissolve to Chan A O Key in
Input On /Off © Dissolve to Chan B 00 Key Out
Ground Patch
© Toggle Dissolve 00 Toggle Key
O Panel © Cut (Dissolve Rate Override)

Active
Fig. 14 -33. Patch panel on HS -200 control console.
to be inserted on the HS -200, and cue it to the start of the sequence by

searching out the appropriate frame number. Use the patch panel ( Fig..
14 -33), and patch from REMOTE VTR jack A to NORMAL and FORWARD (or
to any slow-motion jack, depending on the effect desired). This will cause
the first cue on the tape to start the HS -200 instantly by taking it out of
freeze. The first cue can be either on Editec cue or a 4 -kHz tone -burst cue..
Next, connect the HS-200 playback output to the video input of the vtr,
and play back the tape from the beginning. The "in" cue will start the
HS-200 in playback and permit the output to be viewed on the monitor
of the vtr. The Editec "out" cue will switch the monitor back to vtr output,
providing continuous rehearsal of the entire sequence. If it is satisfactory,
recue the vtr and HS -200 as before, and start the vtr in the Editec record
mode. On this pass, the insert will be recorded exactly as seen in the
rehearsal.
Transitions -The HS -200 can be used to make a cut from one scene
to another or to assist in making a cut. If the total time of the two scenes is
less than 30 seconds, both can be recorded on the disc with precise timing.
The completed sequence of two scenes may then be transferred back to
tape by means of Editec, allowing precise placement of the finished
product.
Cuts may also be achieved on A -B rolls, using two vtr's and an HS -200.
In this case, either the A or B scene may be recorded on the HS -200, and
the other recorded on a vtr. Then, with both signals fed through the auto-
matic dissolver to a second vtr and with playback timings preprogrammed
(by means of Editec and HS -200 cues), the entire sequence can be recorded
in one automatic operation.
Dissolves are accomplished in much the same way at cuts, using the
HS-200 as the source of one signal and a vtr or live camera as the other.
The advantages are that the automatic dissolver in the HS -200 allows
complete control of both placement and length of the dissolve and allows
this operation to be carried out without tying up studio switching equip-
ment. Because the dissolver will dissolve in either direction (from or to the
material on the disc), it is possible to include a number of dissolves on
an A -B roll.
Since a fade is really one -half a dissolve (or a dissolve to or from black) ,
this effect is also readily available with the HS-200. Black is used as the
alternate source to the HS -200 signal, and the fade is programmed for the
rate and placement desired.
Although the patch panel (Fig. 14 -33) can be programmed to control
every function of the HS-200 automatically, either separately or in com-
bination, it is not a complex operation. System functions are enabled
through the patch panel by means of cue pulses, which are directed by
patch cords to the proper inputs. These cue pulses may come from an
outside source such as another HS -200, a video -tape recorder, a manual
control at a remote location, etc., or from within the HS-200 by means of
the memory stores. (It may help to think of these pulses as an extension
of the operator's finger which will push the proper buttons at a preselected
time.) In order to program the system, the cue pulses must be routed from
the source on the FROM side of the patch panel to the proper inputs on the
TO side of the patch panel. In other words, during preprogramming, if the
operation requires selection of two functions together, it also requires a cue
to be directed simultaneously to the same two control inputs on the patch

panel.
Special Effects -There are a number of special effects that can be done
easily on the HS -200, by itself or in combination with other commonly
available equipment. These effects include keying and matting, A -B roll,
time lapse, stop action, titling, and animation.
For producing visual material to be keyed into another picture, the
system can manipulate the frames, one at a time, to record a series of titles
on succeeding frames. Then, with the first title cued in the freeze mode and
th HS -200 signal fed to a keying device, the title can be changed at any
desired rate merely by operating the SINGLE FRAME ADVANCE control.
Assume that it is required to key a product zooming out of a prerecorded
scene. To do this, the zoom could be recorded on the HS -200 (with the
product photographed against black) and both the key and the zoom pre-
programmed by cues from the taped material.
An A -B roll between a vtr and an HS -200, or between two HS- 200's.,
can be programmed by using the patch panel. The A roll, for example, is
assembled on the HS -200. The patch panel is used to cue slow -motion
effects, forward, and reverse on the HS-200, as well as dissolves, cuts.
matting, etc., between the HS -200 and the other HS -200 or the vtr. If a vts
is used for the B roll, a cue is recorded on the cue track at the beginning of
the B roll. This cue is routed to start the HS-200. The A -B roll can be run
over and over again by recuing the HS -200 and the vtr. The cue positions
of all the special effects can be adjusted until the desired effect is achieved.
The A -B roll is then run one more time and recorded onto another vtr.
If a second HS -200 is used for the B roll, the two patch panels are patched
together by using the REMOTE DISC jacks so that when the START PROGRAM
push button on one machine is pressed, it starts both HS- 200's.
With the HS -200, true time -lapse effects are easily produced directly on
video tape. The subject is placed on camera, with the camera orbiter off.
The number of frames desired in each take and the interval between takes
are selected, and the process to be photographed is started. During the
process, the HS-200 is operated at the time interval selected until the
process is finished. Now, when the recording is replayed, the action occurs
in compressed time. To make a time -lapse recording of a biscuit baking,
for example, the total baking time would first be determined. Assume that
it is 8 minutes, and that it is to be reproduced in an 8- second sequence.
If the HS-200 is set to take 1 frame (1/30 second) at each take and make
one take every two seconds during the process, there will be a total of 240
ti
takes, or frames, when the sequence is finished. Playback at the normal
rate of 30 frames per second will take exactly 8 seconds. Due to the flexi-
bility of the HS -200, the playback can be easily adjusted to longer periods
or ',rime. The elapsed time rate generator, an accessory, makes this process
automa !ic. This generator provides time intervals from 1 frame every 300
seconds t',for longer sequences) up to a rate of 30 frames per second. Time
can also be adjusted by varying the number of frames recorded in each

take.
With the HS-200, a stop- action effect can be easily started at any given
cue and then normal playback continued at a later cue. After the scene
is recorded, the placement of a freeze frame is determined by viewing the
action and pressing the ENTER CUE button to store the cued frame. In the
final playback, the FREEZE control is operated to start at the selected points
by means of the patch panel, and after the required interval, the machine is
put into the forward mode by the timer to continue the action. In addition,
the entire playback, including freezes timed to the thirtieth of a second,
may be programmed to play back automatically.
Proper use of the HS -200 single -frame recording capability offers a
source of up to 900 individual titles. Each title is accessible within less
than 4 seconds and may be changed in sequence instantaneously and at any
rate desired. The selected titles may be placed on camera in the desired
order, and a single frame recorded for each. In playback, the HS -200 is
cued to title 1 in the freeze frame mode. As each change of title is required,
the FRAME ADVANCE push button is operated, and the next title appears.
Out-of- sequence titles or frames are located by frame addresses, by means
of the search capability of the recorder. The titles may be incorporated
into the final product as a complete picture or as a "super." With the
keying capability of the HS-200, it is also possible to use the title as a
keyed insert.
The single- or variable -frame record capability of the HS -200 is suited
to any kind of animation currently done on film. It must be remembered,
however, that the actual recording of the individual images ( whether
film or tape) is only one part of the animating process. Even with this
capability, there still remain the problems of manipulating the material to
be animated and the placement of the camera to view the material. For
full cell animation, some adaptation of the film animator's camera stand,
movable "compounds," and peg registration will be necessary. There are,
of course, many useful animations of a simpler nature that can be more
easily achieved. Animated titles are a good example. A title may be made
to appear one letter at a time or one word at a time by starting with a blank
card and adding one increment after each take of the desired number of
frames. A little more complex is the "self- writing" title. The best way of
doing this is to record the complete title first. Then paint or block out a
small portion of lettering at the end of the title and make a 1 -frame record-
ing. Repeat this process, progressing to the beginning of the title, until the
entire word or phrase has been blacked out. Now, if the recording is played
in reverse, the letters will write themselves against the background. Pup-
pets or products can be animated in a similar manner by individual recoru-
ings (of one or more frames) with a small increment or movement be- `
tween successive recordings. The speed of the animated action is, of r ourse,
controlled by the speed of playback, either forward or reverse. '
t
Preplanning -One of the most important operations in any HS -200

recording or editing session is the preplanning of exactly what is required
and exactly how it is to be carried out. After the system operation is under-
stood, it is time- wasting to tie up this valuable equipment when it could
be used for other editing- sequence operations. Preplanning should be done,
as much as possible, at the desk. In other words, develop a procedure for
visualizing the entire operation on paper, making complete lists of cue
times, frame addresses, segment lengths, and sequences before using the
system.
EXERCISES
Q14 -1. In the HS -100 system, how much video material is available on the
disc in the (A) normal record mode and (B) alternate -field record
mode?
Q14 -2. How many disc surfaces are involved in the HS -100 system?
Q14 -3. If a recording made on the video-disc system runs longer than the
maximum capacity of the disc, what happens?
Q14 -4. What is the speed of rotation of the video disc?
Q14 -5. How many lines of video information are recorded during one
revolution of the video disc?
Q14 -6. Does the video head move while it is recording on the video disc?
Q14 -7. What is the major difference between the feeding of record currents
to video -disc heads and the feeding of record currents to video-tape
heads?
Q14 -8. How wide is the recorded track on the video disc?
Q14-9. What is the (A) center -to- center spacing of recorded tracks and
(B) guard band between tracks on the video disc?
Q14 -10. Is speed -up motion available as well as slow motion?
N
APPENDIX A
Table of Prefixes
Power of Ten
Prefix Symbol Definition Multiplier
atto a One millionth of one millionth 10 -18
of one millionth
pico* p One millionth of one millionth 10 -72
micro- µµ One millionth of one millionth 10 -12
micro
nano n One thousandth of one millionth 10 -9
micro µ One millionth 10 -6
milli m One thousandth 10 -3
centi c One hundredth 10 -2
deci d One tenth 10 -1
kilo k One thousand 103
mega M One million 106
giga G One billion 109
tera T One trillion 1012
*Preferred prefix for 10 -12
556
APPENDIX B
Answers to Exercises
CHAPTER 1
Al -1. Remember that in video tape recording, the signal recorded on the tape
is fm. The video signal is pre- emphasized in the modulator and de-
emphasized accordingly in the demodulator to improve the signal-to-
noise ratio. Ahead of the demodulator (where the signal is still
frequency-modulated rf), compensation is provided to allow for tape -
characteristic variations and head variations. Remember in this con-
nection that the highest video frequencies are in the lowest sidebands-
If the response at the higher fm frequencies is increased, lower video
frequencies are increased and highs decreased. If the response at the
lower fm frequencies is increased, lower video frequencies are de-
creased and highs increased.
A1-2. 50/5 = 10 -MHz bandwidth uncompensated. You will find that
modern solid -state video amplifiers, using good high -frequency tran-
sistors (very small capacitances) , use peaking circuits only in isolated
instances.
A1-3. Since
Velocity = Wavelength X Frequency,
then
Velocity
Wavelength =
Frequency
1560 = 156
microinches
10 MHz
A1-4. Twice the head -gap size, or 100 microinches.

Velocity 1560 in/sec
._ 5- Frequency =
Wavelength - 100 fain
15.6 MHz
Al -6. The shelf.

557
A1-7. Since the standard video-to -sync ratio prevails, there is 0.45 volt sync
and 1.05 volts video. Since the modulation sensitivity is set such that
0.7 volt of video causes a deviation of 1.8 MHz, then for 1.05 volts:
0.7 1.05
1.8 x
SO
0.7x = 1.89
x = 2.7 MHz
Then
5 MHz + 2.7 MHz = 7.7 MHz for peak white
What happens here in most low -band recorders is that the shelf is
exceeded, and the modulator "goes to noise." This is a typical overload
condition in which there is a momentary dropout to "noise" for the
duration of the overload (on playback) .
The sync -tip frequency is found as follows:
0.3 - 0.45
0.7 x
so
0.3x = 0.315
x = 1.05 MHz
Then
5 MHz - 1.05 MHz = 3.95 MHz for sync tip
This problem has been solved with the old ratio- and -proportion
technique. A simple check is to reason that if the recorder modulation
sensitivity was set for 2.5 MHz total deviation when the input level
was 1 volt, then the sensitivity is 2.5 MHz /volt. So, if 1.5 volts of
signal is applied, the new deviation is:
(2.5) (1.5) = 3.75 MHz total deviation
This checks the above computations, since:
7.7 MHz peak white
- 3.95 MHz sync tip
3.75 MHz total deviation
CHAPTER 2
A2 -1. Sound track, video track, cue track, control track. NOTE: Modern tape
guides have a radius at the top so that video is no longer recorded on
the audio track at the top of the tape. In early machines, this trans-
verse video track was laid down across the audio track, then removed
by the following audio -track erase head prior to audio recording. The
audio erase head is, of course, still employed to remove any old c
spurious signal component.
A2 -2. No. The tape vacuum guide is moved.
A2 -3. Recording diameter = 0.004 + 2.064 = 2.068 in
Then circumference = 3.14 X 2.068 = 6.493 in

Then velocity = 240 X 6.493 = 1558 in /s
A2 -4. (A) Frequency of H = 15,750 Hz
Then wavelength = 1558/15,750 = 98.92 mils for 63.5/ís
(B) Velocity = 1557 in/s
Then wavelength = 1557/15,750 = 98.85 mils for 63.5 µs
Note that the actual space occupied by one line, or 63.5 µs, is dif-
ferent for different tip projections. It is the space in time, or velocity
vs wavelength of scanning the information, that must remain the same.
A2 -5. Assuming the original recording was made to SMPTE standards, the
playback head was aligned to SMPTE standards, and tape stock has not
stretched, no realignment is required. The reduced head -to-tape veloc-
ity of the playback head directly offsets the difference in tip projection.
Please note that this depends on a drum reference diameter of
exactly 2.064 inches. Remember that there can be a small tolerance in
any dimension. In the same headwheel assembly, this reference diam-
eter is not changing, and if you are recording and playing back only
your own local tapes, there is no problem. But this situation seldom
exists. The guide adjustment is often changed to "match" an outside
tape for optimum performance, particularly in color. Even if everyone
in the business were very critical in standardizing recordings (un-
fortunately, only a dream) , tolerances would still need to be con-
sidered for critical playback.
A2 -6. Note from Fig. 2 -7 that at minimum, the effective diameter must be
2.0658 inches. Assuming the standard drum diameter of 2.064 inches,
there is 2.0658 - 2.064 = 0.0018 inch allowed for tips. Thus, each
tip must be 0.0009 inch, or 0.9 mil, to fall within SMPTE standards.
A recording made with a tip projection of 0.8 mil would likely not
meet the standard radius of rotation.
,v'-7. The vacuum guide is too high or too low (height, or scallop, adjust-
_ ment) This upsets the standard concentricity between head and
.
w.cuum guide.
A2 -8. 1590 ns. See the table in Appendix A.
212 -9. The active picture area is 63.5 µs minus blanking of 11 lts, or 52.5 µs.
So 10 inches corresponds to 52.5 µs, and one inch corresponds to
5.25 .ts. Therefore 1 l.ts is about 1/;-, inch, or between pia and 1/+ inch.
A2 -10. Review Section 2 -5 and Fig. 2 -13B. Departure at the center of the
tape is a measure of actual tip projection.
CHAPTER 3
A3 -1. A frequency -modulated carrier, the reference frequency and deviation

of which depend on whether a low -band monochrome, low -band color,
or high -band mode is used.
A3 -2. A 60 -Hz (nominal) reference pulse derived from the vertical -sync
interval of the signal being recorded (or local sync for local record-
ings).
A3 -3. The head- tachometer signal. This locks the capstan to the velocity
and phase of the headwheel (drum) operation.
A3 -4. The horizontal rate has a reference about every 33 lines. The vertical
reference is 60 Hz from the field rate of the recorded signal.
A3 -5. Same horizontal and vertical rates as in A3 -4. The initial field refer-
ence is usually from local sync, but no feedback information is given
to the capstan or headwheel for frequency and phase lock to local sync.
A3 -6. Same horizontal as above. Feedback information is given to the head
motor and capstan to lock the played -back tape sync to the local
vertical ( field) rate.
A3 -7. The horizontal rate is 15.75 kHz and the vertical rate is 60 Hz
(nominal values), locked in phase to local sync. This is "genlocked"
operation of the vtr.
A3 -8. (A) Line -to -line, or every 63.5 µs. (B) Eight to nine reference pulses
of 0.28 p.s corresponding to the 8 to 9 cycles of color -subcarrier burst,
once each line on the back porch of horizontal sync
A3 -9. Older machines and modern monochrome -only systems still use the
control -track and head -tachometer comparison of Fig. 3 -6. Later
systems use the basic method of Fig. 3 -8, with refinements. More
information on this is given in Chapter 7.
A3 -10. Velocity lock, then phase lock.
CHAPTER 4
A4 -1. (A) Master erase, audio erase, audio record /play, audio simulplay.
(B) Master erase, control -track record /play, cue erase, cue recor
play, control-track simulplay.
A4 -2. The position of the vacuum guide.
A4 -3. The capstan only. The supply and take -up reel motors merely provide
constant torque.
A4 -4. To take up excess tape if any speed variation occurs due to changes in
operating modes, and to stop the machine automatically if the tape
breaks or a reel runs out of tape.
A4 -5. To lift the tape from contact with the master erase head in either the
forward or reverse wind mode. This minimizes tape wear by eliminat-
ing contact during any operation not requiring contact with the mat«
erase head.
A4 -6. Air guides, tape lifter, and vacuum guide.
A4 -7. Headwheel air bearing and vacuum-guide air cylinder.
A4 -8. The one nearest to the +70 -volt trigger, in this case setup.
CHAPTER 5
A5 -1. Only if the heads are within plus or minus 0.005 µs (5 ns) error.
This is within plus or minus 1.5 seconds of arc (approximately).
The latest high -band heads (also used for low band) meet this speci-
fication, and no electronic quadrature correctors are employed with
these heads. In practice, even when this tolerance is slightly exceeded,
built -in time -correction circuitry removes the error (Chapter 6) .
A5 -2. (A) 1 to 8 MHz
(B) 1to16MHz
A5 -3. (A) 4.28 MHz (H) 7.9 MHz
(B) 5.0 MHz (I) 10 MHz
(C) 6.8 MHz
(D) 5.5 MHz
(E) 5.79 MHz
(F) 6.5 MHz
(G) 7.06 MHz
A5 -4. Irregularities in tape coating that cause video loss, or "white flashes,"
for the duration of the dropout. This may result in tiny white dots or
much larger flashes, depending on the area of irregularity.
A5 -5. (A) Approximately 100 V (p -p) (B) Approximately 5 mV (p -p) .
.
A5 -6. The headwheel (drum) tachometer signal.

AS -7. (A) From the modulator. (B) From the fm switcher.
A5 -8. To eliminate noise which would cause amplitude variations of the fm
component.
A5 -9. So that switching is timed during the blanking (actually sync) inter-
val. (In the Ampex system, switching is timed to occur just ahead of
horizontal sync.)
A5 -10. No. It is used to compensate the region of resonance (around 10
MHz).
CHAPTER 6
A6 -1. Time error.

A6 -2. A voltage -controlled video delay line.
A6 -3. The voltage- controlled video delay line is made to have a time function
opposite to the error being corrected.
A6 -4. Tape sync supplies both the sampling pulse and the sawtooth or
trapezoid. A dc error is obtained by making a phase comparison be-
tween the edge of tape sync and a reference sync pulse generated by a
long- time-constant afc loop locked to tape sync.
A6 -5. Same as in A6 -4 except that the reference sync is obtained from the
station generator instead of tape sync.
A6 -6. 1000 nanoseconds (1 microsecond) .

A6 -7. Reverse biased.
A6 -8. Tape burst and local subcarrier reference.
CHAPTER 7
A7 -1. Yes.
A7 -2. The control-track record current is increased until a just perceptible
inflection (shoulder) occurs in the region of the zero -axis crossing as
observed on the output of the simulplay control -track head.
A7 -3. Program audio head to assure that the audio signal is being recorded,
cue -track audio head to monitor the cue-track audio while recording,
and control-track head to assure that the control track is being laid
down and to permit proper adjustment of the record current.
A7 -4. You don't. This is determined by the arrangement shown in Fig. 7 -1.
A7 -5. By the TRACKING (CONTROL TRACK PHASE) knob, which adjusts
the delay in the control -track playback signal.
A7-6. Velocity lock followed by phase lock of the headwheel (drum) , from
which a trapezoid signal is formed. Then this trapezoid is sampled
with a pulse from:
1. Reference vertical from local sync.
2. Reference vertical from comparison of local vertical sync with
demodulated tape vertical sync.
3. Final step: Same as (2) but added "time- modulated" vertical
comparison to horizontal -sync errors.
A7 -7. (A) Reference frame pulse is compared in phase with pulse obtained
by dividing frequency of control-track signal by eight; reference frame
pulse is compared with tape frame pulse. (B) Vertical framing. (C)
Because the servos must be able to lock in to the closest horizontal
pulse. Therefore, the framing pulse must be dropped as a reference.
A7-8. (A) A 30 -pps signal recorded along with the control -track signal,
indicating occurrence of field 1. (B) Same as A. (C) A 30 -pps signal
derived from the vertical interval of the local sync generator. (D) A
30 -pps signal derived from the vertical interval of the demodulated
tape playback signal.
A7 -9. Yes. The tight control of time base requires timing (not necessarily
phasing) to the local sync generator.
A7 -10. To obtain greater sensitivity in microseconds per volt of error signal.
CHAPTER 8
A8-1. A reduction in rf amplitude, and therefore a video loss. "White
flashes" occur in the picture for the duration of the dropout. Drop-
outs are caused by irregularities in the tape coating, either flaws in the
tape itself or dust and other foreign matter. They result in small white
dots or much larger flashes depending on the area of the irregularity.
A8 -2. The rf level from the recorded tape signal.
A8 -3. Not if the machine playing back to the dubbing system contains a
dropout compensation system. However, if it does not, any dub
(second -generation tape) will contain the same dropouts, and any
subsequent dub (third generation) cannot be replayed without the
dropouts even though a DOC is used. That is, the DOC acts only on
rf signal levels from the tape being played, and not on any "recorded -
in" dropout in the video signal.
A8 -4. The video from the previously scanned line is substituted. This is
possible because successive lines (or several lines) contain a large
amount of redundant information.
A8 -5. Since the substituted signal is delayed one line, the color phase is
shifted by 180 °. Therefore, luminance and chrominance information
are separated, and the chroma is phase- inverted to obtain the correct
color polarity.
CHAPTER 9
A9 -1. (1) Hue shifts. (2) Saturation variations.

A9 -2. (A) Velocity error. (B) Differences in head -to -head frequency re-
sponse, tape surface variations, and tape -to -head contact differences.
A9 -3. To the average value of the previous 24 revolutions (100 milliseconds)
of the headwheel.
A9 -4. The analog circuits produce the output signals which correct the
errors. The digital circuits supply to the analog circuits the timing
and control pulses required for line -to -line correction.
A9 -5. Since the ATC error signal is applied to the variable delay line at
the start of each video line and remains constant throughout the line,
the error waveform appears as a series of steps. This restores video
timing to normal at the start of each line, but it allows errors to
accumulate along the line.
A9 -6. It generates voltage ramps which connect the beginning and end of
each step of the signal described in A9 -5, to achieve continuous cor-
a rection of the signal throughout the line.
A9 -7. The VEC analog circuits measure the difference between successive step
voltages (beginning of one line to beginning of next line, up or down
in voltage) and store the difference voltage (positive or negative) in
a capacitor memory.
A9 -8. Because generation of a ramp that will connect successive ATC steps
requires advance knowledge of the slope.
A9 -9. The X signals correspond to lines, and the Y signals correspond to

heads. One X signal is provided for each horizontal line in a head
scanning period, and one Y signal is provided for each of the four
head scan periods.
A9 -10. No. The readout and write -in of the CAC error for a particular line
occur during the same line.
CHAPTER 10
A10 -1. (A) The new scene cannot be synchronous to the existing program
on the tape, and this causes servo upset, at both the start and stop of
the insert, until new lockup is achieved. (B) The distance between
the record and erase heads would cause a double (and nonsynchro-
nous) recording to be made for approximately 9 inches of tape.
410 -2. (A) The insert mode retains the already existing control -track signal
on the tape during the record mode. The new recording starts at the
desired place and ends before the old recording ends. Thus, it requires
an ingoing and an outgoing splice. (B) The add -on (assemble) mode
requires only an ingoing splice because it simply adds on to the
original recording. A new control track is recorded as in normal
recording. Synchronism with the original recording is maintained by
making the capstan servo a slave to a capstan memory circuit, which
has measured and stored the speed of the original recording.
410 -3. Improper phasing of the tachometer signal with respect to the origi-
nal recorded signal.
410 -4. (A) The original control -track signal. (B) A dc voltage from the
capstan memory circuit (controlling a vco).
410 -5. A computer system which enables the operator to "practice" a com-
plete editing procedure by watching a picture monitor (and listen-
ing to an audio monitor) without disturbing the original recording.
Then, when the entire procedure is satisfactory, the information is
stored again in the computer, and a single push button initiates auto-
matic editing.
CHAPTER 11
A11 -1. Start with midrange. The purpose of this control is to permit setting
the individual channel equalizers for equalization of all chroma (and
burst) levels. If any of the individual controls is nearing the limit
of its range, change the master equalizer accordingly.
A11 -2. In the latest systems, playback levels are controlled by agc. The con- t
trols marked MGC (manual gains) are effective only on test-signal
positions. Obviously, in an older tape system, you must first equalize
all levels (rf) as indicated on the cro at the switcher monitoring
position.
Incidentally, the best way to adjust these levels (for color) is to

look at a color monitor at the system output and adjust for minimum
transients between individual bands of 16 lines. Then adjust the
equalizers. Go back and forth between these two adjustments.
A11 -3. Depending on the severity of this difference, you will see some
banding in red and blue. Slightly trim the damping "R" adjustment
on the playback amplifiers to equalize the "bottoming" levels (most
recent high -band systems only) .
A11 -4. Review Section 11 -2 (subsection "Adjusting Monochrome and Color
ATC Error Gains ") and Fig. 11 -17.
A11 -5. First of all, never adjust the demodulator gain until you have first
checked frequency and deviation in the E -E circuit mode. These
adjustments affect the output level of the demodulator. Then a stan-
dard tape will play back at the same level from the demodulator with
properly equalized individual playback amplifiers ahead of this de-
modulation. If the level is low, the tape was underdeviated; if the
level is high, the tape was overdeviated.
A11 -6. First, this point should be made: The best way to adjust demodulator
balance is to put the system in the monochrome mode and adjust
balance for minimum rf in sync and blanking. (Some color filters
[when in the color mode) do not permit adjustment in this manner.)
Then on playback, balance should remain the same provided the tape
was recorded on the proper frequency.
Another reason for lack of balance between the E -E mode and
actual playback is excessive levels from the individual playback
amplifiers ahead of the demodulator. This can produce an over-
load condition leading to unbalance. Reduce these levels and note
the effect on the demodulator output.
Incidentally, you cannot see rf in the superblack region when the
system is in the color mode. You normally adjust balance in the
monochrome mode and trust this is retained for color. You can dou-
ble -check this by observing a color monitor at the system output and
adjusting the balance of the demodulator for minimum moiré on
color bars.
A11 -7. 0.714 volt for video, 0.286 volt for sync.
A11 -8. fe = 012 = 5 MHz

A11 -9. (A) Tilt of bar when on the proper graticule time base (shift
centering when necessary)
(B) Pulse height, width, and overshoots when on proper time base
(C) K- factor of pulse-to -bar amplitude ratio
A11 -10. Always make the burst phase correct first, so that the chroma side-
band information is correct relative to burst. Then adjust the system
phase to match the standard color -bar generator at the output of the
special- effects amplifier.
A11 -11. By measuring the playback performance characteristic with the sine -
squared and window signals on the SMPTE standard alignment tape.
Then compare a local recording on the same playback. On modern
high -band systems, the performance should be better than that of the
monochrome standard alignment tape. The overall K- factor should
be well within 4 percent for the T -pulse and 2 percent for the 2T-
pulse.
A11 -12. From 1.8 MHz to 15 MHz (double -sideband rf signal) .
A11 -13. In the external -sync mode, local external reference sync from the
sync generator is added to the video signal. In the regenerated -sync
mode, the processing system forms new composite sync.
A11 -14. Yes. The RCA test tape is recorded on a 1.0331 -inch guide. Strictly
speaking, this would cause some difference between new and worn
heads. In practical everyday use, this is negligible, and you need have
no hesitation in setting the guide for any head (new or worn if more
than 1 mil of tip remains) against the standard tape.
A11 -15. 1 mil. Some users still employ a gauge to measure tip protrusion,
but
this is not advisable due to possible damage to a good head. The use
of the standard multiburst signal from a test tape (as described in
this chapter) is most satisfactory. The back -away method (to find the
point of departure) with precision measurement of the headwheel
radius is sometimes (but rarely) used.
A11 -16. No. When you record at a lower -than- standard penetration, tape
contact is lost as the head wears, and the result is about the same life
as with standard penetration. In addition, interchangeability is lost.
A11 -17. Here are the things that must be done: All sources to be lapped must
be synchronous, so the tape machine (s) must be operated in the
Pixlock (RCA) or automatic (Ampex) mode (tightest servo re-
quirement). Then each tape system must be phased horizontally for
the same timing at the leading edge of sync. Then the tape system
phase must be adjusted for proper burst phase at the mixing point.
It is also quite desirable that the blanking widths be approximately
the same, that the front -porch and back -porch positions be the same,
and that the positions of the two bursts on the back porch be very
close.
A11 -18. Some high -band tapes have a tendency to stick in the transport, and
this is normally the fault of the tape. High-band tapes have a very
smooth surface. A smooth surface lying against another smooth sur-
face has more of a tendency to cling to the guide, posts, etc., than a
rough surface against a smooth surface would have. This effect is
termed tape seizure or tape adhesion. The drag is actually so great
under this condition that capstan slippage occurs. If it happens often,
check all transport tensions, including the capstan pressure roller,
and thoroughly degauss the entire tape path.
CHAPTER 12
Al2 -1. Axial alignment and axial position. Review Section 12 -2.
412 -2. "Tip ringing," a fault of the head itself, caused by deterioration of the
magnetic -gap material.
Al2 -3. Feed a multiburst signal to the system input. Response in the E -E
mode is immediately apparent by noting the difference between the
video output and the video input, with the machine in the stop
mode. Playback characteristics are judged by use of the standard test
tape. Recording characteristics then will be apparent by making a
recording of the local multiburst signal and playing this back on the
same and other machines of the common installation. The latter
step is simply to ascertain the degree of likeness or difference of all
machines.
Al2 -4. (A) 0.125H /cm X 25.
(B) 0.25H /cm X 25.
412 -5. The modulated 20T- pulse.
Al2 -6. Displacement of colors from the proper luminance position.
Al2 -7. Bent supply or take -up reels, improper tape- tension adjustments
(always refer to specific instruction book for system), dirty heads or
tape path, worn heads, contaminated capstan shaft or pinch roller,
insufficient capstan -to -tape pressure, bad tape.
Al2 -8. Magnetized head, dirty head, worn head, poor tape.
CHAPTER 13
A13 -1. Yes, if the tape is wound onto a conventional reel. The same type of
head and the same recording standards prevail for both the cassette/
cartridge and reel -to -reel systems.
A13 -2. Yes, if the tape length does not exceed the capacity of the cassette or
cartridge.
A13 -3. It can be purchased either way. If shared electronics is used, the other
machine must be an Ampex AVR -1.
A13 -4. It can be purchased either way. If shared electronics is used, the other
machine must be an RCA TR -60 or TR -70.
A13 -5. (A) Start of tape. (B) End of tape.
A13 -6. (A) A hole in the tape (sensed by a photocell) (B) A reflective
.
tape (sensed by a photocell).

413 -7. (A) Two. (B) Two. (C) One. (D) Two.
413 -8. Two seconds maximum.
CHAPTER 14
A14 -1. (A) 30 seconds. (B) 60 seconds.
A14 -2. Four, the top and bottom surfaces of the top and bottom discs (two -
disc, four -head system).
A14 -3. The latest 30 seconds (60 seconds for alternate -field recording) is
maintained in storage. Prior recorded material is progressively
erased.
A14 -4. 60 revolutions per second (3600 rpm) .
A14 -5. 262% lines (one field), beginning and ending during the vertical -
blanking interval.
414-6. No. It is held stationary while the disc makes a complete revolution.
When the next head starts recording, the head that was recording
previously is stepped to a new position.
A14 -7. For conventional tape recording, the record currents are fed continu-
ously to all four rotating heads, and a separate head does the erasing.
In video disc recording, the record currents are gated on and off to
each head, and the head used for recording a track is also used to
erase that track by gating the erase current to the head at the proper
time.
A14 -8. 7 mils.
A14 -9. (A) 10 mils. (B) 3 mils.
A14 -10. Yes, continuously variable up to twice normal speed.
Index
Add -on Automatic

operation, 296-298, 301 -307 color circuitry, use of in maintenance,
recording, speed control during, 300 419-422
Adder, sync, 182 control panel, 454-456
Address, 263 mode, 195 -197
Afc Time Correction, 153
loop, 177 -179 Automation, sequential, 471
modulator, 111 -113 Axial position
sync pulses, 179 check for, 394 -396
Air error, 47 -48
bearings, 48 Azimuth
headwheel, supply for, 84 check for, 394
cylinder, vacuum -guide, supply for, error, 47
84
filter, 84 -85 Bandwidth, 18
guide, 77 capstan servo, 431
systems, 78 -85 Bias
tank control, 82 -84 control adjustment, 445 -446
Alternate-field high- frequency, 73
logic, 531 Blanking, 179
playback, 495 -496 Blowers, cooling, 84 -85
recording, 491 Burst
reproduce, 514 phase, 361 -362
Ampex signal, keyed, 399
ACR -25, 451 -459
HS -100 video recorder /reproducer, CAC, 255
484 -487 analog system, 273
HS -200 teleproduction system, digital system, 277
539 -555 Capstan servo, 465 -468
system design considerations, near -lock detector, 435 -436
459 -468 RCA, 213 -223
VR- 2000B, measurement in, 414 -415 Carriage
Amplifier, playback, 128 -131 error -correction logic, 539
Amtec, 72, 153 logic
Analog system Al module, 537
CAC, 273 A2 module, 535 -536
VEC, 261 -263 retiming logic, 537 -539
Animation, 539 -540, 545 reversing, 490 -491
Assemble, 298, 550 -551 logic, 536
Assembly, 541 -542 stepping sequence in reverse motion,
ATC, 72, 153 536 -537
color, 162 -170 Cartridge, 469 -470
error gains, adjusting, 360 system, 451
monochrome, 158 -162 Cassette, 459
principles of, 153 -158 system, 451
Audio CATC, 153
system, 72 -75 Cavec, 253, 255 -259
test, 443 -444 Channel switching, electronic, 131 -145
track, original, 299 -300 Character generator, BE 400, 327 -328
569
Chroma Cue -cont

amplitude corrector, 271 -281 track, original, 299 -300
inverter Cuts, flash, 544 -545
logic, 509, 531 -532 Cutting, 542
module, 507 -514
phase, 493 -494 Degaussing, 363 -364
inverter, 507 -509 Delay line, 503 -505
Clamp pulse, 180 electronically variable, 158 -160
Cleaning, 363 Demodulation basics, 95 -102
Clipping, selective frequency, 338 Demodulator, 507
Clock circuitry, modern, 145 -150
generator, 526 -527 module, 502 -503
MDA, 521 -522 Departure point, 371
Closed -loop operation, 275 -277 Detector
Coarse switch, 4 X 2, 133 -137 adjustments, near -lock, 438
Code, edit, 314 -316 control- track, 432 -435
Color Diameter, recording, 30
ATC, 72, 153, 162 -170 Differential
error gain, adjusting, 360 gain measurements, 407 -408
Automatic Time Correction, 153 phase measurements, 407 -408
bars, use of, 355 -360 Digital system
circuitry, automatic, use of in main- CAC, 277
tenance, 419 -422 VEC, 264
error correction systems, advanced, Diode, voltage -variable -capacitance, 106
need for, 253 -255 Disc
Time Error Correction, 153 rotation, 488
Colortec, 72, 153 servo unit, 486
Concentricity, 37 logic, 532 -539
Console Dissolves, automated, of variable
control, 453 -454 length, 541
system, 539 Dissolving, 542
transport, 453 Distortion
Control (s ) geometric, 155
air -tank, 82 -84 test, 446 -447
circuits DNEB, 160
sequential programming, 541 DOC; see also dropout compensator
transport, 85 -88 module, 245 -249
console, 453 -454 system, RCA, 237 -249
dropout compensator, 245 DPEB, 160
functions, 514 -522 Drop -frame technique, 314
logic system, 514 -539 Dropout(s), 34, 236
modes, servo, 195 -201 compensator, 236
operating, 545 -549 Ampex, 249 -252
panel controls, 245
automatic, 454 -456 processor module, 240 -245
manual, 456 vertical, 240
secondary, 457 Drum, 27
signal, 457 Dsu, 486
play, 475 Dual
processing logic, 521 heterodyne modulation, 104
record, 477 speed transport, 29 -30
track
detector, 432 -435
frame pulse, 65, 214 E-E
record and simulplay waveforms, circuit, 91
223 -225 frequency response, 447 -448
sensor adjustments, 438 Edit
transport, 476 code, 314 -316
unit, 487 generator, BE 520, 316 -324
Cross talk, sync, 397 -398 control system, 310
Cue pulse, 65, 214
control circuit, 521 time code reader, BE 420, 324 -327
INDEX 571
Editing Field identification pulse, 522

electronic, programmed, principles of, Filter (s)
309 -314 air, 84 -85
system, EECO, 314 -331 Bode, 146
Editor Flash cuts, 544 -545
electronic, programmer for BE 210, Fm
328 -331 equalization, electronically controlled,
time -code 277 -281
RCA, 310 -312 reference, 113 -121
using, 312 4 X 2 coarse switch, 133 -137
Elapsed time rate generator, 553 Frame
Electronic (s) advance control, 521
Editor, 288 pulse, 214
programmer, BE 210, 328 -331 random access to, 540
signal unit, 486 -487 Freeze
logic, 522 -532 frame
splicing, 387 -389 or field, 540
Ampex, 288 -295 time controlled, 540
RCA, 295 -309 mode, 493
system, 477 -481 Frequency
-to- electronic circuit, 91 clipping, selective, 338
Emergency operations, 381 -386 modulation, 19
Equalization range for television tape, required,
curve, 15 -16 16 -19
fm, electronically controlled, 277 -281 response
playback, setting, 357 -360 E -E, 447 -448
Equalizer, rf, 502 modulation- demodulation, 121 -125
Erase record /playback, 448 -449
head
tuning, 444 -445 Gain
video, 293 -bandwidth product, 17 -18
test feature, 300 -301 differential, measurement of, 407 -408
Error(s) error, ATC, adjusting, 360
axial -position, 47 -48 video, remote, 182 -185
azimuth, 47 Gap, head, 14
gains, ATC, adjusting, 360 Gate drivers, 505
geometric, 32 Gated sync separator, 177
guide height, 354 Gating tubes, 99
penetration, 354 Generation of timing signals, 264 -267
quadrature, 46 -47 Generator
time - character, BE 400, 327 -328
base, 70 edit-code, BE 520, 316 -324
space, 37 -48 elapsed time rate, 553
timing, limitation of, 192 sync, local, optimizing, 396-398
velocity, 33 timing, 138 -145
Esu, 486 -487 Guide height error, 354
Exhaust air system, 79 -82
Extract, 263
Half -
banding, 48, 395 -396
Fablock, 430, 431 -432 line delay logic, 530-531
maintenance procedure for, 439 -441 Head (s )
setup procedure, 437 -438 alignment procedures, 365 -375
subboards A and B, 437 assembly, video, 394 -396
Fading, 542 clogged, cleaning, 394
Fast erase, video, 293
forward logic, 528 -530
control, 519 -521 motor, speed of, 64
motion, 544 movement, 489
reverse control, 519 -521 resonance test, signal flow for, 131
search, 498 retiming logic, 530
logic module, 525 -526 servo, 465 -468
Head (s) -cont L azy

-switch pulse generator, 145 Servo Simulator, 206
switching sequence in reverse motion, Simulated Servo, 229
536 Leader, spot, 458 -459
-to -tape velocity, 30 -33 Level checks, 364 -365
video Lifter, tape, 82
identification of, 53 -54 Line
optimizing, 375 -379 interlace, 493
Headwheel, 27, 52 sync pulse, 522
air bearing, supply for, 84 Linelock, 206, 212 -213, 220 -221
assembly, 48 -55 playback mode, 229
servo, RCA, 201 -213 Lip sync, 73
unlock indicator, 225 -229 Lobes, pulse, 340 -341
Heterodyne color technique, 104 Logic
High - alternate -field, 531
band test tapes, 351 -355 carriage
frequency error correction, 539
control adjustment, 445 -446 retiming, 537 -539
problem, solving, 21 -23 reversing, 536
pressure air system, 82 -84 chroma- inverter, 509, 531 -532
Horizontal circuitry
adjustment, 367 -372 applications of, 430-441
mode, 197 -199 servicing of, 430 -441
phasing provisions, 300 control processing, 521
HS -100 disc servo unit, 532 -539
functional description of, 487 -498 electronics signal unit, 522 -532
video half -line delay, 530 -531
recorder /reproducer, Ampex, head, 528 -530
484 -487 retiming, 530
signal system, 498 -514 push- button, 516 -519
HS -200 teleproduction system, Ampex, slow-motion, 528
539 -555 system, control, 514 -539
HW unlock indicator, 225 -229 Low -
band
color, deviation for, 103 -104
Indexing- editing "Mini- Modules," 316 test tapes, 344
Indicator (s) , 545 -549 frequency problem, solving, 19 -21
headwheel unlock, 225 -229 pressure air and vacuum system,
status, 331 79-82
Insert, 551 -552
operation, 296-298, 307 -309
Insertion, 542
Interchange S- distortion, 41 -42 Maintenance
Interlace, line, 493 audio system, 441 -449
Internal reference circuitry, 172 -175 automatic color circuitry used in,
Intersync, 186 419 -422
servo system, basics of, 189 -201 panel, 458
Inverter procedure for Fablock, 439 -441
logic, chroma, 509 Manual control panel, 456
module, chroma, 507-514 MATC, 153
phase, chroma, 507 -509 Matting, 541, 543
Memory
Jitter, 382 -383 capacitance, 257
Jog(s), 354 control pulses, generation of, 267 -271
adjustment, 367 -372 panel, 549
"Mini- Modules," 316
K -factor, 338, 404 -405 Mode selection, 298 -299
system, 341 Modulation
k factor, 18 circuitry, 106 -121
Keyboard time entry, 329 -demodulation
Keyed -burst signal, 399 frequency response, 121 -125
Keying, 541, 543 systems, refinements in, 102 -106
INDEX 573
Modulation-cont Phase-cont
index, 20 chroma, 493-494
process, basics of, 92 -95 differential, measurement of, 407 -408
Modulator inverter, chroma, 507 -509
afc, 111 -113 system, 361 -362
module, 498 Phasing, horizontal, 300
30-MHz, 503 Pinch roller, 78
Monitoring, built-in, 541 Pixlock, 206, 207 -212
Monochrome ATC, 158 -162 achieved circuit, 212
error gain, adjusting, 360 playback mode, 228 -229
Multiburst amplitude levels, 338 Platch circuit, 212 -213
Play
Near -lock, 432 control, 475
detector mode, 206
adjustments, 438 functions, servo, 194 -195
capstan servo, 435 -436 Playback
NEB, 158 alternate -field, 495 -496
Noise amplifier, 128-131
null adjustments, 444 -445 checks before, 380
positional, 162 equalization, setting, 357 -360
Nonphased color mode, 169 mode, 65 -69
Nonsynchronous signals, playback and Linelock, 229
recording of, 418 Pixlock, 228 -229
Normal Switchlock, 228
mode, 200 tonewheel, 228
speed, 491 -493 operation, 217 -223
Npc mode; see nonphased color mode synchronous, 69 -72
Numerical display panel, 545 -546 variable -speed, 541
Positional noise, 162
Open-loop operation, 273 -275 Preamplifier, video, 125 -128
Operation, system, 549 -555 Preplanning, 555
Optimizing Preset mode, 201
local sync generator, 396 -398 Processed video, 180 -182
video heads, 375 -379 Processing
Opu, 487 pulse, 177
Orientation of recorded signals, 27 -30 signal, 175 -182
Oscillator, slow- motion, 521 unit, output, 487
Oscilloscope method of measuring S/N video, 180
ratio, 409 -410 Processor module, dropout, 240 -245
Output Programmed electronic editing, princi-
processing unit, 487 ples of, 309 -314
video module, 507 Programmer, electronic editor, BE 210,
Overdeviation, 376 328 -331
Programming control circuits, sequen-
Pad, video, 400 -401 tial, 541
Panel Pulse (s )
automatic, control, 454 -456 clamp, 180
maintenance, 458 edit, 65, 214
manual control, 456 field identification, 522
memory, 549 frame, 214
numerical display, 545 -546 control- track, 65
record, 546 -549 generator, head-switch, 145
secondary control, 457 line sync, 522
setup, 458 memory -control, generation of,
signal control, 457 267 -271
Peaking circuits, 401 processing, 177
PEB, 158 servo reference, 522
Penetration error, 354 start, 177
Performance, overall, evaluation of, sync, afc, 179
391 -393 20T modulated, 405 -406
Phase vertical reference, 525
burst, 361 -362 Push -button logic, 516 -519
Quadrature Ringing, 343 -344

adjustment, 373 -375 rip, 396
error, 46 -47 Rolls, A and B, 312 -313, 542 -543
Quadruplex tape system, recording stan-
dards for, 23 -24 5- distortion,interchange, 41 -42
Quantizer, slow- motion, 527 -528 Safety margins, 392
Sample- and -hold circuit, 263
checking, 428 -430
Random access to any frame, 540 Sawtooth signal, 407
RCA TCR -100, 468 -481 Scallop adjustment, 365 -367
Reader, edit time code, BE 420, 324 -327 Secondary control panel, 457
Recommended practice RP10, SMPTE, Selective frequency clipping, 338
344 -347 Self -check procedures, 365 -375
Record Sequence control, 329
amplifier module, 498 -499 Servo (s)
control, 477 capstan, 465 -468
mode, 203 -206, 227 near -lock detector for, 435 -436
functions, servo, 193 -194 RCA, 213 -223
normal, 214 -215 control modes, 195 -201
panel, 546 -549 functions, understanding, 424 -428
-playback frequency response, head, 465 -468
448 -449 headwheel, RCA, 201 -213
process, 488 -491 play -mode functions, 194-195
signal path, 491 record mode, 64-65
standards, 113 -121 functions, 193 -194
waveform, control- track, 223 -225 reference pulse, 522
Recording stability, checking, 422 -424
alternate -field, 491 system
checks during, 379 -380 Intersync, basics of, 189 -201
mode, 60 -65 relationships in, 186-189
standards for quadruplex tape system, terminology, 153
23 -24 unit, disc, 486
Reference vacuum -guide, 229 -235
circuitry, internal, 172 -175 Setup
fm, 113 -121 mode, 227
frame pulse, 214 panel, 458
Regulator, -5 volt, 437 procedure, preliminary, 363 -364
Rehearsal, 541 Shoe servo, 229
Remote Sidebands, 20
control, two -way, 541 Signal (s)
video gain, 182 -185 control panel, 457
Reproduce flow
alternate -field, 514 head -resonance test, 131
forward, 509 -512 normal, 128 -131
process, 491 -498 rf level detection, 131
reverse, 512 -514 path
signal path, 496 -498 record, 491
Resolution, 16 -17 reproduce, 496 -498
Resonance, head, signal flow for test of, processing, 175 -182
131 recorded, orientation of, 27 -30
Reverse system, 462 -465
motion, 495, 545 video, HS -100, 498 -514
carriage stepping sequence in, test, use of, 399 -408
536 -537 -to -noise ratio, 408 -418
command K, 536 measurement, 447
head switching sequence in, 536 oscilloscope method of measuring,
reproduce, 512 -514 409 -410
Rf RCA TR -70, measurement in,
copy operation, video channel, 416-418
141 -143 vtvm method of measuring,
level detection, signal flow for, 131 409-410
switcher module, 499 unit, electronics, 486 -487
INDEX 575
Simulplay waveform, control- track, Sync -cont

223 -225 cross talk, 397 -398
Sin"- window signal, 338 generator, local, optimizing, 396 -398
Single -disc, two -head video system, processor, tape, 170 -172
482 -484 pulses, afc, 179
Skew selection, 179 -180
adjustment, 367 -372 separator
effect, 40 gated, 177
Skewing, 354 module, 522
Slow motion, 494-495 Synchroguide circuitry, 397
forward, 544 Synchronizer, limited range, 329 -331
logic, 528 Synchronous playback, 69 -72
oscillator, 521 System concept in tape operations,
quantizer, 527 -528 334 -362
SMPTE
de- emphasis curve, 121 pulse, 338
T-
edit code, 314 Tachometer, 62

Recommended Practice RP10, Tape
344 -347 frame pulse, 214
Sound, sweetening, 313 -314 horizontal alignment, 71
Special effects, 553 -554 new, dropouts in, 392
Speed control, add -on recording, during, operations, system concept in,
300 334 -362
Splice(s) production terminology, 332 -334
formation of, 296 speed, control of, 300
modes, 221 -223 sync processor, 170 -172
types of, 295 -296 systems, 25
Splicing television, frequency range required
electronic, 387 -389 for, 16 -19
Ampex, 288 -295 test
RCA, 295 -309 high band, 351 -355
mechanical, 283 -287 low band, 344
Spot leader, 458 -459 use of, 347 -351
Stability, servo, checking, 422 -424 transport, 58 -60, 76 -78
Stairstep signal, 407 vertical alignment, 70
Standardizations, 334 TCR -100, RCA, 468 -481
Standards Teleproduction system, Ampex HS -200,
international, 182 539 -555
record, 113 -121 Television signal, elements of, 190
Standby mode, 227 Terminology, tape production, 332 -334
Start pulse, 177 Test
Status indicators, 331 audio, 443 -444
Still- frame, 493 distortion, 446 -447
Stop action, 544 erase, 300 -301
Superimposure, 543 setup, 443
Sweetening sound, 313 -314 signals, use of, 399 -408
Switch tapes
coarse, 4 X 2, 133 -137 high band, 351 -355
equalizer and, sync -channel fm, 138 low band, 344
video -channel fm, 137 -138 use of, 347 -351
write, 262 Tha, 155
Switcher module, rf, 499 30 -MHz
Switching delay-line switcher and demodulator,
channel, electronic, 131 -145 505 -507
gate, 505 modulator, 503
logic module, 526 Time
Switchlock, 207 -base
playback mode, 228 correcting devices, 72
Sync error, 70
adder, 182 code
channel, 143 -145 edit, reader for, BE 420, 324 -327
fm switch and equalizer, 138 editor, RCA, 310 -312

Television Broadcasting Tape and Disc Recording Systems Ennes

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Television Broadcasting Tape and Disc Recording Systems Ennes

Uploaded by

Copyright:

Available Formats

M Tape and Dis

HOWARD W. SAMS & CO., INC.

Copyright © 1973 by Howard W. Sams & Co., Inc., Indianapolis,

All rights reserved. Reproduction or use, without express permis-

International Standard Book Number: 0- 672 -20933 -0

Broadcasting: Equipment, Systems, and Operating Fundamentals, and Tele-

BASIC CONCEPTS IN QUADRUPLEX VIDEO TAPE SYSTEMS . 13

1 -1. The Time -Space Relationship 13

THE QUADRUPLEX HEAD: THE BASIC TIME BASE . 27

2 -1. Orientation of Recorded Signals 27

BASIC REQUIREMENTS: WHAT THE SYSTEM MUST Do . . 57

3 -1. The Tape Transport 58

3 -3. The Playback Mode 65

THE TAPE TRANSPORT, AIR SYSTEMS, AND CONTROL CIRCUITRY . 76

THE VIDEO PROCESSING SYSTEM 151

6 -7. Signal Processing 175

THE SERVO SYSTEM 186

THE TAPE -DROPOUT COMPENSATION SYSTEM . 236

ADVANCED COLOR -ERROR CORRECTION SYSTEMS . 253

SPLICING AND EDITING FOR QUADRUPLEX RECORDING 283

10 -4. Principles of Programmed Electronic Editing . . . 309

QUADRUPLEX TAPE SYSTEM OPERATIONS . 332

QUADRUPLEX TAPE SYSTEM MAINTENANCE 391

QUADRUPLEX -TAPE CASSETTE /CARTRIDGE SYSTEMS 451

VIDEO DISC RECORDING SYSTEMS . . . . 482

TABLE OF PREFIXES 556

ANSWERS TO EXERCISES 557

Basic Concepts in Quadruplex

Television tape recording requires a magnetic -field manipulation that

1 -1. THE TIME -SPACE RELATIONSHIP

Although the time -space relationship is important to all communication

inch). At a tape speed of 15 in /s, the recorded wavelength of a 1000 -Hz

Fig. 1 -1. Fundamental time -space

Very Low Frequency, Small Rate of Change Across Gap

High Frequency With Wavelength Equal to Gap Size -

Fig. 1 -2. Relationship between head -gap size and frequency.

speed on frequency response.

Frequency (Log Scale)

frequency response suffers. (There is a very small rate of change across

Response Limited by Losses

Frequency (Log Scale) Frequency (Log Scale)

(A) Uncompensated response. (B) Typical equalization curve.

1 -2. REQUIRED FREQUENCY RANGE FOR TELEVISION TAPE

Fig. 1 -5. Wedges for interpreting

Fig. 1 -6. Scanning beam encountering sharp vertical line.

Fig. 1 -7. Wedges for interpreting

Table 1 -1 relates bandwidth to rise time and horizontal -resolving power in

Table 1 -1. Relationship of Bandwidth, Rise Time, and

Bandwidth (MHz) Rise Time (µs) TV Lines

1 -3. SOLVING THE LOW- FREQUENCY PROBLEM

with a carrier frequency of 5 MHz clamped to the signal- blanking level.

IMPORTANT NOTE: The frequencies and deviation shown in Fig. 1 -8

apply only to low-band monochrome standards. The discussion in this

With the carrier frequency clamped at the video-blanking level, it is

Lowest sideband for Maximum Sync Tip I Blanking Peak

I-0. 3V-r- 0.7 V -

'Charles E. Anderson, "Signal Translation Through the Ampex Videotape

1 -4. SOLVING THE HIGH -FREQUENCY PROBLEM

speed of 7.5 in /s, the recorded wavelength is equal to:

Rotating Head Assembly

Tape Pulled Toward

Fig. 1 -9. Parallel feed of four rotating heads (simplified diagram).

I -- --Af 2.5 MHz -I Af 3 MHz -- {